Abstract
Introduction:
Recent research has begun to demonstrate high rates of poor behavioral health outcomes among homeless former foster, but with relatively little attention in the scientific literature. Because social networks have been shown to impact behavioral health outcomes, we need to better understand the network characteristics of homeless young adults with a history of foster care.
Methods:
Data were collected from 184 homeless former foster youth at a drop-in center in Hollywood, California. A series of logistic regressions were conducted for network engagement behaviors that may be impacted by foster care experiences.
Results:
Youth were largely African American, Latino, and Mixed race, approximately 22 years old, and majority male. Significant results from the logistic regressions indicated that youth experiencing homelessness for the first time before exiting foster care were more likely to have alters engaging in condomless sex, and youth with higher numbers of foster care placements were more likely to have alters engaging in methamphetamine use.
Conclusion:
These results suggest that certain foster care experiences impact the types of alters with which youths engage. Time spent in placement was significantly associated with alter behaviors, evident from homeless experiences and multiple foster care placements, negatively impacting the types of alters with which youths are connected, and thusly the risk factors for the youth themselves. Frequent network disruptions may be associated with engagement with risky alters. Included are intervention-based implications for providers as well as structural-based implications for child welfare systems.
Keywords: social network analysis, social network engagement, homeless youth, foster youth, HIV, risk behaviors, emerging adulthood
Introduction
Recent research has begun to demonstrate high rates of poor behavioral health outcomes among homeless former foster youth (Courtney et al., 2011; Hudson & Nandy, 2012; Nyamathi et al., 2012; Yoshioka-Maxwell & Rice, 2016). However, given the number risk factors common to many homeless former foster youth, this population has received relatively little attention in the scientific literature, as compared with homeless youth and youth with foster care experiences. Furthermore, because social networks have been shown to impact behavioral health outcomes for both homeless youth and former foster youth (Bao, Whitbeck, & Hoyt, 2000; Ennett, Baily, & Federman, 1990; Halkitis et al., 2013; Joseph, O’Connor, Briskman, Maughan, & Scott, 2014; Kidd, 2003; Rice, Milburn, & Rotheram-Borus, 2007; Rice & Rhoades, 2013; Rice, Stein, & Milburn, 2008; Wenzel, Hsu, Zhou, & Tucker, 2012) and little research has been conducted on the networks of homeless former foster youth, there is a need to better understand the network characteristics of homeless young adults with a history of foster care and the impact of these characteristics on behavioral health outcomes for this at-risk population.
Background
An Overview of Foster Youth and Former Foster Youth
There are nearly 428,000 youth in foster care on any given day in the United States, with 670,000 youth spending time in foster care in 2015 (U.S. Department of Health and Human Services, 2008). Foster youth and former foster youth are a subpopulation of the homeless populace who are at a high risk for factors such as unemployment, low educational attainment, early parenthood, criminal activity, and mental health conditions (Daining & DePanfilis, 2007). Childhood abuse is associated with high rates of substance use and substance abuse disorders, sexual activity at a younger age, increased risk of certain sexually transmitted illnesses (STIs), and multiple sex partners, making former foster youth at high risk of experiencing poor behavioral health outcomes, including HIV and STIs (Ahrens et al., 2010; Benjet, Borges, Medina-Mora, & Méndez, 2012; Black, Oberlander, & Lewis, 2009; Kerr et al., 2009; Schilling, Aseltine, & Gore, 2007). Additionally, rates of substance use and risky sex practices among older foster youth are often higher than those for the general public (Braciszewski & Stout, 2012; Carpenter, Clyman, Davidson, & Steiner, 2001; DiClemente, Crittenden, & Rose, 2008; Vaughn, Ollie, McMillen, Scott, & Munson, 2007; White, O’Brien, White, Pecora, & Phillips, 2008).
Life experiences unique to former foster youth may affect their HIV-risk behaviors. Approximately 40% of homeless youth in Los Angeles report having been placed in foster care at some point (Yoshioka-Maxwell et al., 2015). Because these youths have been involved in the foster care system, they are eligible for a wide array of services that other homeless youth are not. Despite these advantages, some former foster youth have higher levels of HIV-risk behaviors, including engagement in exchange sex, vaginal intercourse at an earlier age, and more sex partners (Ahrens et al., 2010; Hudson & Nandy, 2012; Nyamathi, Hudson, Greengold, & Leake, 2012b).
An Overview of Youth Homelessness and Accompanying Risk Factors
As of 2017, in the Unites States, one in 10 young adults aged 18–25 and one in 30 youth aged 13–27 has experienced homelessness over the course of a year (Morton et al., 2017). This homelessness is a risk factor for a variety of negative outcomes, but rates of drug use are particularly high for homeless youth, correlating to use at an early age, polysubstance use, use within the last 30 days, and likelihood of lifetime use. Homelessness also serves as a risk factor for the presence of a substance abuse disorder (Bousman et al., 2005; Greene, Ennett, & Ringwalt, 1997; Kipke, Montgomery, Simon, & Iversom, 1997; Salomonson-Sautel et al., 2008). Moreover, drug use among homeless youth carries more risks than just the associated poor health outcomes, such as sexual risk-taking, particularly where combined heroin/stimulant users are concerned (Gleghorn, Marx, Vittinghoff, & Katz, 1998).
This sexual risk-taking behavior is not merely related to drug use. These behaviors from this population also include participation in exchange sex, or exchanging sex for a good or service. Like drug use, exchange sex correlates to additional risks, including victimization, substance use, suicide attempts, and STIs (Greene et al., 1997). The rates of exchange sex are higher for homeless youth than for youth who are not homeless, and these high rates increase when a substance use history is present (Chettiar, Shannon, Wood, Zhang, & Kerr, 2010; Greene et al., 1997; Kral, Molnar, Booth, & Watters, 1997).
Research has indicated that youth exiting foster care are at greater risk for experiencing homelessness than their peers without a history of foster care (Dworsky, Napolitano, & Courtney, 2013), and homeless adults are 8 times as likely to have a history of foster care than the housed general public (Reilly, 2003). Many homeless youths leave home to escape abuse; a recent study demonstrated that up to 85% of homeless youths have experienced either physical or sexual abuse before becoming homeless, with 42% experiencing both (Keeshin & Campbell, 2011). However, despite the attempts of these youth to escape abuse situations, homelessness puts them at a higher risk for further victimization, with evidence demonstrating that homeless youth experience disproportionately high rates of robbery, assault, and sexual assault (Thrane, Hoyt, Whitbeck, & Yoder, 2006). The accumulation of risk factors common to both homeless youth and former foster youth place homeless former foster youth in the nexus of life experiences that carry an extreme risk of engaging in HIV-risk behaviors, necessitating an increase in research to address the unique needs of this population. Foster youth continue to face these risk factors after emancipation and well into adulthood, but research has not examined behavioral health outcomes faced by former foster youth who have also experienced homelessness.
Theoretical Background
Peers play an important role in the lives of homeless young adults (Whitbeck & Hoyt, 1999). As these young adults move away from family-centered networks and toward peer-centered networks, homeless youth become highly influenced by their street-based peers (Furman & Buhrmeter, 1992; Rice, Milburn, Rotheram-Borus, Mallett, & Rosenthal, 2005; Rice, Stein, & Milburn, 2008). Theories of social networks and risk amplification offer models for explaining these peer influences and the impact they may have on individuals’ behavioral health.
One of the central tenets of Social Network Theory posits that similarity breeds connection (Brechwald & Prinstein, 2011; McPherson, Smith-Lovin, & Cook, 2001). Consequently, peer behavior has been shown to be strongly correlated with individual behavior (Weis & Hawkins, 1981). Network characteristics may inform us about the types of behavior in which an individual is likely to engage (Rice, Tulbert, Cederbaum, Barman-Adhikari, & Milburn, 2012). Thus, in examining the connections and behavioral health patterns among similar individuals, researchers may gain a clearer understanding of the rates and prevalence of these behaviors in the greater network. The Risk Amplification and Abatement Model (RAAM) views negative contact with socializing agents as amplifying risk while positive contact with socializing agents as abating it (Milburn et al., 2009). This model is an extension of the Risk Amplification Model (RAM), which focuses solely on the negative outcomes that occur from negative experiences (Cauce, Paradise, Embry, Morgan, Lohr, Theofelis, … & Wagner, 1998). Whitbeck and Hoyt (1999) used the RAAM to show how parental problems and family abuse increase a youth’s likelihood of engaging with deviant social networks. Findings from the development of this model indicated that both time spent with deviant networks and time spent alone impact levels of engagement in drug risk behaviors through the continued reinforcement of engagement in risk behaviors from the negative contact with socializing agents. RAAM is a useful model for the population of homeless youth because it attempts to explain the impact of negative life events and negative developmental trajectories while considering the effect that positive contact has on abating risk. In this perspective, negative contact with socializing agents, such as deviant social networks, amplifies risk of engaging in risk-taking behaviors.
Studies examining the social networks of homeless young adults indicate that support networks play a critical role in mitigating the negative effects of homelessness on mental and behavioral health. Negative network ties have been shown to increase anti-social behavior, depression, risk-taking behaviors, engagement in drug risk behaviors, and perceptions of negative social support (Bao, Whitbeck, & Hoyt, 2000; Ennett, Baily, & Federman, 1990; Halkitis et al., 2013; Kidd, 2003; Rice, Milburn, & Rotheram-Borus, 2007; Rice & Rhoades, 2013; Rice et al., 2008; Wenzel, Hsu, Zhou, & Tucker, 2012).
Within the context of the foster care experience, this model can be slightly expanded to explain why former foster youth may uniquely engage in certain levels of risk-taking behaviors. Given the experiences common to many former foster youths, including abuse, instability, and institutionalization, this model may help predict a former foster youth’s likelihood of engagement in deviant social networks. Therefore, because homeless former foster youth share common experiences, they may be more likely to have shared network characteristics, which, in turn, impact their engagement in risky drug use behaviors (figure 1).
Figure 1.
Analytic Model of the Relationship between Foster Care and Network Engagement
For young people with a history of foster care, connectedness is particularly impactful on wellbeing, in that, engagement in quality relationships is associated with fewer disruptive behaviors (Joseph, O’Connor, Briskman, Maughan, & Scott, 2014). In contrast, fewer social ties are associated with additional risk factors and increased network disruption for those with a history of foster care compared with other young adult populations (Perry, 2006). Whereas networks are important for youth during transitional periods, young adults from foster care tend to struggle with maintaining relationships with their birth families and have a difficult time adjusting to the often-abrupt changes that come with transitions and discharge (Barth, 1990; Blakeslee, 2011; Collins, Spencer, & Ward, 2010). Furthermore, former foster youth experiencing homelessness additionally tend to lack support during transition out of foster care and be isolated from family, friends, and other support networks, increasing their risk for poor outcomes (De La Haye et al., 2012; Fischer & Breakey, 1991).
Research on the impact life histories on social network engagement is still in its infancy and requires further examination. Risk factors common to homeless youth and former foster youth are well-established, but examination of the social networks of these youth is limited. Generally, social networks have been shown to affect behavioral health outcomes among homeless youth and former foster youth, such as drug and sex risk-taking behaviors (Bao et al., 2000; Ennett et al., 1990; Halkitis et al., 2013; Joseph et al., 2014; Kidd, 2003; Rice et al., 2007; Rice & Rhoades, 2013; Rice et al., 2008; Wenzel et al., 2012). Although networks are associated with risk-taking behaviors of foster youth and homeless youth, with the exception of one study (Yoshioka-Maxwell, Rice, Rhoades, & Winetrobe, 2015), little research to date has examined the role of networks for youth with histories of both homelessness and foster care. Outside the context of homelessness, research has established that for former foster youth, connectedness affects well-being, in that, engagement in quality relationships is associated with fewer health and behavioral problems (Joseph et al., 2014). Although family contact and family support have been associated with resilience (Jones, 2012, 2013), this population experiences fewer ties and faces more network disruption compared with other young adult populations (Perry, 2006). Furthermore, long-term foster care and the process of discharge have been shown to hinder the development of an ideal support structure by increasing network disruption, which results in sparse social networks and negatively affects emotional, relational, and behavioral health (Blakeslee, 2011; Collins, 2004; Courtney et al., 2005; Geenen & Powers, 2007; Perry, 2006; Reilly, 2003). Although networks are important for youth during transitional periods, young adults from foster care tend to struggle to maintain relationships with birth families and attachments to supportive adults and have a difficult time adjusting to the often-abrupt changes that come with transitions and discharge from foster care (Barth, 1990; Berzin, 2008; Blakeslee, 2011; Reilly, 2003).
Furthermore, negative network ties have been shown have a causal impact on increased HIV-risk behaviors among homeless youth in general, regardless of foster care history (Bao et al., 2000; Ennett et al., 1990; Halkitis et al., 2013; Kidd, 2003; Milburn et al., 2009; Rice et al., 2007; Rice, Milburn, Rotheram-Borus, Mallett, & Rosenthal, 2005; Rice & Rhoades, 2013; Rice et al., 2008; Wenzel et al., 2012). Studies examining the social networks of homeless young adults indicate that support networks play a critical role in mitigating the negative effects of homelessness on mental and behavioral health. Negative network ties have been shown to increase anti-social behavior, depression, risk-taking behaviors, engagement in drug risk behaviors, and perceptions of negative social support (Bao et al., 2000; Ennett et al., 1990; Halkitis et al., 2013; Kidd, 2003; Rice et al., 2007; Rice & Rhoades, 2013; Rice et al., 2008; Wenzel et al., 2012). Finally, one analysis examining the sociometric network structure of homeless former foster youth indicated that former foster youth occupy a unique space within this network, remaining largely along the periphery, and that this space impacts their engagement in recent and lifetime substance use, which is important information for determining appropriate interventions (Yoshioka-Maxwell & Rice, 2016). These results imply that aspects unique to the foster care subpopulation influence youth affiliation and network structure, indicating a need for further examination of the factors predicting the formation of these affiliations and this structure.
It has been established that techniques such as Social Network Analysis can be an important tool in the science of social work. Rice and Yoshioka-Maxwell (2015) outlined the importance of Social Network Analysis within the field of social work, as an essential mechanism for tying together human experiences with social change. However, studies examining these behaviors have been limited in the data collected on former foster youth and have only made gross comparisons between youth with and without foster care experiences. These studies have not delved into how the heterogeneity of foster care experiences affects the risk-taking, housing stability, and peer-involvement trajectories of youth with and without experiences of foster care.
Additional data sampling and analyses should be conducted to determine which foster care experiences impact social network engagement, with the expectation that experiences unique to the foster care system significantly impact individuals’ social network ties, including the type of ties and behaviors of those ties. Understanding the pathways between foster care experiences and network engagement may provide information required to understand the types of social network engagement by former foster youth that increase their likelihood of engaging in risky networks and behaviors that can potentially impact other behavioral health outcomes. Based on the current knowledge of behavioral health outcomes and the importance of network engagement for former foster youth, this analysis seeks to better understand network characteristics of homeless former foster youth and how these characteristics are associated with behavioral health outcomes such as risk-taking behaviors.
Methods
Sampling
Data were collected from 184 homeless former foster youth at a drop-in center in Hollywood, California using the risk-behavior questionnaire modeled after the YouthNet Study (Rice, 2012), which broadly examined social networks and addressed mental and behavioral health issues among homeless youth in Los Angeles. Data for the current analysis was collected during 2- to 4-week intervals over three periods (two summers and one winter) from 2015 to 2016. Foster care experiences were measured through the Foster Care Experiences Assessment, a qualitative assessment created by Yoshioka-Maxwell in 2014 to gather information on the range of experiences among former foster youth both during and after placement. This qualitative survey, conducted through semi-structured interviews with 20 homeless former foster youth, was later coded and turned into a quantitative survey used for quantitative analysis of foster care experience.
All clients receiving services at the respective agency during data-collection periods were eligible for screening. Inclusion criteria included receipt of services by the drop-in center as well as a history of out-of-home care with the child welfare system. All clients receiving services through the drop-in center were already considered homeless as defined by McKinney-Vento (McKinney-Vento Homeless Assistance Act of 1988), per the policy of the drop-in center. Recruitment was conducted for approximately 2–4 weeks; during that time, recruiters were present at the agency to approach youth for the duration of service provision hours. Youth new to the agency first completed the agency’s intake process before beginning the study to ensure they met the eligibility requirements for the agency (and, thus, the study). A consistent set of two research staff members were responsible for all recruitment to prevent youth completing the survey multiple times within each data-collection period.
Signed voluntary informed consent was obtained from each youth, with the caveats that child abuse and suicidal and homicidal intentions would be reported. Informed consent was obtained from youth 18 years and older, and informed assent was obtained from youth 14 to 17 years old. The Institutional Review Board (IRB) at the University of Southern California waived parental consent, as homeless youth younger than 18 years are unaccompanied minors who may not have a parent or adult guardian who could give consent. Interviewers received approximately 40 hours of training, including lectures, role-playing, mock surveys, ethics training, and training in emergency procedures. The study consisted of two parts: a computerized self-administered survey and a social network interview. Two phases of data collection were included to gather two types of data. In the self-administered questionnaire, behavioral health and foster care experience questions were asked of the participants. As social network survey was separately administered to gather information on the types of connections youth maintained as well as the behaviors of individuals in their networks. The two data sets were linked to better examine the relationship between social networks and behavioral health outcomes and foster care experiences.
Part 1: Online, self-administered questionnaire.
Participants privately entered answers into the computer as they read questions on the computer screen or listened to the questions being read to them. Preprogrammed skip patterns advanced participants to the next appropriate question after a response was entered. These methods reduced nonresponse rates to sensitive questions about potentially socially undesirable activities, such as sexual behaviors, illicit substance use behaviors, and criminal activity (Ghanem, Hutton, Zenilman, Zimba, & Erbelding, 2005; Jones, 2003; Macalino, Celentano, Latkin, Strathdee, & Vlahov, 2002; Metzger et al., 2000; Turner et al., 1998).
Part 2: Network assessment.
After the youth indicated relevant members of their social networks, questions about types of relationships and attributes of each nomination (i.e., alter) were asked. The same name generator developed for YouthNet (Rice, 2012) was used. Name generators involved questions designed to elicit the naming of relevant alters along some specified criterion (Laumann, Marsden, & Prensky, 1991). Respondents were given a prompt defining a criterion, such as “Who do you turn to for advice or support?” In the YouthNet study, extensive qualitative work was conducted before study implementation to create questions to elicit the nomination of significant network ties among homeless youth. The resultant name generator includes 17 prompts related to the following: people you talk to, hang out/kick it/chill with; people you have sex with or hook up with; people you party with or drink or use drugs with; old friends from home; people you talk to (on the phone, by email); people from where you are staying (squatting with); people you see at this agency; and other people you know on the streets. For the purpose of this analysis, egocentric network data was collected to measure the types and behaviors of ties between individuals, rather than to examine whole-network structural patterns (sociometric network analysis).
Measures
Sociodemographic variables.
Age, race, gender, and sexual orientation were obtained through self-report measures. Age was calculated by coding the youths’ reported birth dates into an age based on the date of their interview. Racial categories included American Indian/Alaska Native, Asian, Black/African American, Native Hawaiian or other API, White, Latino/Hispanic, and mixed race. Due to the low numbers of American Indian/Alaska Native, Asian, and Native Hawaiian or other API, those categories were coded into an “other” category. For the purpose of the logistic regressions performed in this analysis, race was further dichotomized into “Black/African American” and “all other races” due to the frequency distribution of the race categories (Black/African American=1, All other races=0). Questions pertaining to gender included the following options: male, female, transgender (male to female), and transgender (female to male). Due to the low number of transgender participants, all responses related to gender were coded into male or female, depending on the gender with which participants identified (male=1, fmale=0). Variables pertaining to sexual orientation included homosexual, queer, bisexual, heterosexual, and questioning/unsure. Due to the response rates across the orientations, sexual orientation was coded into “heterosexual” and “LGBTQ” (heterosexual=1, LGBTQ=0). All reference categories were selected solely on the basis of frequency.
Foster care variables.
A number of foster care experience variables were chosen to describe basic experiences in foster care placements. Age at first foster care placement was measured on a 7-point scale, ranging from placement at birth to placement at age 17 (1=at birth, 2= 1 year or younger; 3= 2–3 years old, 4= 4–6 years old, 5= 7–10 years old, 6= 11–13, 7=14–17 years old). Time spent in placement was measured through a 6-point scale, ranging from less than a year to 15 or more years. Age at exit from foster care was measured on a 4-point scale and included the categories of 5 years or younger, 6–11 years old, 12–17 years old, and under 18 years old generally. Housing situation after transitioning out of foster care included a number of options for housing such as family, family of origin, adoptive family, transitional living facility, couch surfing, homelessness, independent living, shelter, jail, rehab, and foster family. Number of foster care placements was measured through a 5-point scale, ranging from one or two placements to 20 or more placements. Type of placement included kinship care, foster care, group home, juvenile detention, psychiatric hospital, and camp. These options were not mutually exclusive but meant to capture the range of placements that a person may have throughout their childhood. Reason for placement included physical abuse, sexual abuse, neglect, parental drug problems, truancy, suicide attempt, personal drug use, parental psychiatric problems, placement at birth, and other. Finally, general feelings regarding foster care were obtained. Feelings of being supported and feelings of being respected were measured through a 5-point scale that included “never,” “almost never,” “sometimes,” “almost always,” and “always.” For the purpose of the logistic regressions, the questions pertaining to feelings about placement were coded into three categories: low, medium, and high levels of support and respect. All foster care variables selected were chosen based on their importance in previous literature or the frequency of discussion in the qualitative interviews.
Homelessness variables.
Basic information regarding individuals’ homelessness experiences including overall time spent homeless measured in months and years, and age at first homelessness, measured in participant age. Timing of homelessness was asked to gauge timing of homelessness either before leaving foster care (before age 5, between 6–11 years old, between 12–18 years old) or after leaving foster care (as a minor, as an adult). Perceived cause of homelessness included: aging out of foster care, self-blame, disagreements with family/friends, kicked out of their house, lost a job/need a job, evicted, “lost my roommate and couldn’t pay for rent:, drug use, “I stopped trying”, no foster care resources, no support system, by choice, “I made poor choices”, needed transitional services, family problems from childhood. Time currently spent homeless was measures through a 6-point scale, ranging from less than a year to 14–18 years. Time spent homeless over the lifetime was measured through a 6-point scale ranging from less than a year to 11–13 years.
HIV-risk behaviors.
HIV-risk variables were selected from the CDC’s Youth Risk Behavior Survey (YRBS) (Brenner, Collins, Kahn, Warren, & Williams, 1995), which is used to track youth risk behavior and has been tested for validity and reliability. Measures that were included in final models included a series of dichotomized variables such as ever having sex, condom use at last sexual encounter, drug use with sex at last sexual encounter, exchange sex over the lifetime, exchange sex over the last 3 months, condom use with exchange sex, injection drug use over the lifetime, injection drug use over the last 3 months, and ever having an STI test (other than HIV). HIV testing behavior was measured through a 3-point scale, within the last 3 months, 3–6 months ago, 6 or more months ago.
Egocentric network variables.
Egocentric network variables were taken from the network survey answered by the youth regarding their social networks. Measures of alter centrality, network density, and average total ties were created from the egocentric network data to provide context to the larger social network. Two types of network-based variables were used for this analysis. First, variables regarding the types of ties present in individuals’ networks were used, including ties from home, foster care, a group home, kinship care, a partner, a friend, or staff and ties that youth talk to about sex, or felt they could confide in. This first set of ties includes the presence and name/acronym of individuals who fit these categories. Ties from home are those individuals considered to be connections from the participant’s last home, or place they most consider home. Similarly ties from foster care, group home, and/or kinship care include any individual they know from these locations. Similarly, individuals considered to be partners, friends, or staff were left for the participant to define. Second, variables regarding the behaviors of individuals’ ties were used, including ties they believe to engage in condomless sex, who object to condom use, use methamphetamines, use heroine, use cocaine, or use injection drugs. The responses to these questions were subjective, but participants were instructed to select individuals with whom they have discussed or seen these behaviors. All of these network ties were recorded and used as dichotomized variables, with the exception of centrality, density, and average total ties, which were used as continuous data.
Analyses
One of the main objectives of this analysis was to explore some of the unique characteristics of homeless former foster youth and their network engagement. Because the descriptive statistics of homeless former foster youth were previously established (Yoshioka-Maxwell, Rice, 2019), for this analysis, demographics for this population were only provided for foster care demographics. Beyond that, a chi-square analysis was run to first explore the impact of alter foster carte status with alter type and alter behavior to determine whether or not type and alter behavior was a function of alter foster care status. Once this was established, a series of multivariate logistic regressions were conducted for network engagement behaviors that may be impacted by foster care experiences. Logistic regression was chosen because of the interest in the presence or absence of an alter behavior or type, rather than a count of the number of these connections. Additionally, most alters named between two or three alters, so the variability does not exist to create a count. These models were built through the use of variables previously established as impactful for this population (Yoshioka-Maxwell, Rice, Rhoades, Winetrobe, 2015; Yoshioka-Maxwell, Rice, 2019). For the foster care experience models, individual models were tested for time spent in foster care, number of foster care placements, foster care exit age, and “first homelessness experience before foster care” with types and behaviors of ties the youth indicated. These models controlled for age, race, gender, sexual orientation, feelings of support in foster care, and feelings of respect in foster care. All foster care variables were chosen based on previous literature’s establishment of risk and protective factors associated with specific experiences, such as time spent in foster care, number of foster care placements, and age at exit from foster care. All analyses were conducted using SAS 9.3.
Results
Demographics
Basic demographic statistics were run for sociodemographic, foster care, homelessness, and HIV-risk behavior variables. As demonstrated in Table 1, the average age of the sample was 21.11 years (SD=1.95), with the majority of youth reporting their race as Black/African American (47.37%), followed by Mixed race (21.05%). White and Latino youth represented 12.87% of the sample independently. The majority of youth reported being male (67.63%) and heterosexual (70.41%), with 15.38% reporting their sexual orientation as bisexual, and 8.28% as homosexual.
Table 1.
Basic sample demographics (n=173).
| All Youth (n=173) | |||
|---|---|---|---|
| n (%) | Mean | SD | |
| Age | 21.99 | 1.95 | |
| Race | |||
| American Indian/Alaska Native | 8 (4.68) | ||
| Asian | 1 (.58) | ||
| Black or African American | 81 (47.37) | ||
| Native Hawaiian or other API | 1 (.58) | ||
| White | 22 (12.87) | ||
| Latino/Hispanic | 22 (12.87) | ||
| Mixed Race | 36 (21.05) | ||
| Gender | |||
| Male | 117 (67.63) | ||
| Female | 48 (27.75) | ||
| Transgender- male to female | 7 (4.05) | ||
| Transgender- female to male | 1 (.58) | ||
| Sexual Orientation | |||
| Homosexual | 14 (8.28) | ||
| Queer | 3 (1.78) | ||
| Bisexual | 26 (15.38) | ||
| Heterosexual | 119 (70.41) | ||
| Questioning/Unsure | 7 (4.14) | ||
Where foster care demographics were concerned, table 2a depicts that the majority of youth reported being placed in foster care between 14–17 years old (21.47%), 2–3 years old (20.25%) or 11–13 years old (17.79%), while reports of time spent in the foster care system covered a broad time span, with 22.98% reporting being in placement for 15 or more years, and 20.50% reporting being placed for less than a year. Over half of the participants (56.43%) reported transitioning out of care before the age of 18, with the next highest group being 12–17 years old (26.43%), with 33.54% of the sample reporting 1–2 placements, and 22.36% reporting 3–4 placements. Still, 16.15% of the sample reported 10 or more placements, and 15.53% reported 20 or more placements. The majority of youth reported spending time in foster homes (65.41%), with equal percentages of youth reporting placements in kinship care and group homes (13.21%).
Table 2a.
Basic foster care demographics (n=173)
| All Youth (n=173) | |
|---|---|
| n (%) | |
| Age at Placement | |
| At birth | 23 (14.11) |
| Under 1 | 9 (5.52) |
| 2–3 years old | 33 (20.25) |
| 4–6 years old | 15 (9.20) |
| 7–10 years old | 19 (11.66) |
| 11–13 years old | 29 (17.79) |
| 14–17 years old | 35 (21.47) |
| Time Spent in Placement | |
| Less than 1 year | 33 (20.50) |
| 2–4 years | 32 (19.88) |
| 5–7 years | 35 (21.74) |
| 8–10 years | 13 (8.07) |
| 11–14 years | 11 (6.83) |
| 15 or more years | 37 (22.98) |
| Age at Exit from Placement | |
| Under 18 years old | 79 (56.43) |
| 12–17 years old | 37 (26.43) |
| 6–11 years old | 13 (9.29) |
| 5 years old or younger | 11 (7.86) |
| Housing after FC Transition | |
| Family | 26 (16.25) |
| Family of origin | 28 (17.50) |
| Family/adoptive | 13 (8.13) |
| Transitional living facility | 19 (11.88) |
| Couch surfing | 10 (6.25) |
| Homeless | 28 (17.50) |
| Independent living | 5 (3.13) |
| Shelter | 6 (3.75) |
| Jail | 10 (6.25) |
| Rehab | 2 (1.25) |
| Foster family | 13 (8.13) |
Table 2b demonstrates that youth reported neglect (38.15%), physical abuse (30.06%), and parental drug problems (31.79%) as the reasons for placement in foster care. Housing situation immediately following transition from foster care varied widely, with 16.25% reporting living with family members, 17.50% reporting living with their family of origin, and 17.50% reported immediate homelessness. When asked about their feelings regarding foster care, the majority of youth reported feeling always (29.19%) or almost always (25.47%) supported, and always (26.25%) or almost always (26.88%) respected.
Table 2b.
Basic foster care demographics (n=173)
| All Youth (n=173) | |
|---|---|
| n (%) | |
| Number of FC Placements | |
| 1 to 2 | 54 (33.54) |
| 3 to 4 | 36 (22.36) |
| 5 to 9 | 20 (12.42) |
| 10+ | 26 (16.15) |
| 20+ | 25 (15.53) |
| Type of Placement | |
| Kinship | 21 (13.21) |
| Foster home | 104 (65.41) |
| Group home | 21 (13.21) |
| Juvenile detention | 5 (3.14) |
| Emergency shelter | 6 (3.77) |
| Psychiatric hospital | 1 (.63) |
| Camp | 1 (.63) |
| Placement Reason | |
| Physical abuse | 52 (30.06) |
| Sexual abuse | 25 (14.45) |
| Neglect | 66 (38.15) |
| Parental drug problems | 55 (31.79) |
| Truancy | 18 (10.40) |
| Suicide attempt | 5 (2.89) |
| Personal drug use | 16 (9.25) |
| Parental psychiatric problems | 29 (16.76) |
| Placed at birth | 20 (11.56) |
| Other | 23 (13.29) |
| Feelings about Foster Care | |
| Support | |
| Never | 23 (14.29) |
| Almost never | 23 (14.29) |
| Sometimes | 27 (16.77) |
| Almost always | 41 (25.47) |
| Always | 47 (29.19) |
| Respect | |
| Never | 25 (15.63) |
| Almost never | 28 (17.50) |
| Sometimes | 22 (13.75) |
| Almost always | 43 (26.88) |
| Always | 42 (26.25) |
Where homelessness demographics are concerned, Table 3 demonstrates that on average, youth had been homeless for 2.27 years (SD=2.74), with their age at first homeless experience being 16.52 years. 73.33% of youth considered themselves “homeless”. Where first homeless experiences were concerned, 37.75% reported first becoming homeless after leaving foster care, and as an adult, while 30.46% reported their first homeless experience being before they left foster care, age 12–18. They reported their causes of homelessness largely as a result of aging out of the foster care system (29.14%), followed by self-blame (19.21%), disagreements with family/friends (13.25%), and being kicked out (13.91%). The majority of youth reported that they had currently been homeless for less than a year (38.67%), and that their time spent homeless over the course of their lifetime was 3–4 years (26.80%), followed by 5–7 years (22.88%).
Table 3.
Homelessness characteristics among former foster youth (=173)
| All Youth (n=173) | |||
|---|---|---|---|
| n (%) | Mean | SD | |
| Time homeless | 2.27 | 2.74 | |
| Age at first homelessness | 16.52 | 4 | |
| Do you consider yourself homeless | 121 (73.33) | ||
| First homeless experience | |||
| Before leaving FC- before 5 years old | 11 (7.28) | ||
| Before leaving FC- 6–11 years old | 15 (9.93) | ||
| Before leaving FC- 12–18 years old | 46 (30.46) | ||
| After leaving FC- as a minor | 22 (14.57) | ||
| After leaving FC- as an adult | 57 (37.75) | ||
| Cause of your homelessness | |||
| Aged out of foster care | 44 (29.14) | ||
| I blame myself | 29 (19.21) | ||
| Disagreements with family/friends | 20 (13.25) | ||
| Kicked out | 21 (13.91) | ||
| Lost a job/need a job | 5 (3.31) | ||
| Evicted | 3 (1.99) | ||
| Lost my roommate and cant pay rent | 4 (2.65) | ||
| Drugs | 1 (.66) | ||
| I stopped trying | 2 (1.32) | ||
| No FC resources | 4 (2.65) | ||
| No support system | 4 (2.65) | ||
| By choice | 1 (.66) | ||
| I made poor choices | 2 (1.32) | ||
| Needed transitional services | 4 (2.65) | ||
| Family problems from childhood | 3 (1.99) | ||
| Time homeless- current | |||
| Less than 1 year | 58 (38.67) | ||
| 1–2 years | 31 (20.67) | ||
| 3–4 years | 32 (21.33) | ||
| 5–7 years | 23 (15.33) | ||
| 8–10 years | 1 (.67) | ||
| 11–13 years | 3 (2.00) | ||
| 14–18 years | 2 (1.33) | ||
| Time homeless- lifetime | |||
| Less than 1 year | 30 (19.21) | ||
| 1–2 years | 28 (18.30) | ||
| 3–4 years | 41 (26.80) | ||
| 5–7 years | 35 (22.88) | ||
| 8–10 years | 6 (3.92) | ||
| 11–13 years | 4 (2.61) | ||
Finally, where HIV-risk behavior is concerned, table 4 demonstrates that 90.00% reported ever having sex, with just under half (47.71%) having used a condom at their last sexual encounter. Of youth that reported ever having sex, 41.18% reported using drugs with sex at their last sexual encounter. For exchange sex, 26.14% reported ever engaging in exchange sex, with 50.00% of those participants reporting having engaged in exchange sex recently and 37.21% reported using a condom with exchange sex. Of the full sample, 90.75% reported ever having an HIV test, with 67.52% of those youth reporting having had an HIV test in the last 3 months, and 24.84% reporting testing 3–6 months ago.
Table 4.
Sex risk variables among homeless former foster youth (n=173)
| All Youth (n=173) | |
|---|---|
| n (%) | |
| Ever had sex | 153 (90.00) |
| Condom use- last sex | 73 (47.71) |
| Drug use- last sex | 63 (41.18) |
| Exchange sex- lifetime | 40 (26.14) |
| Recent exchange sex | 22 (50.00) |
| Condom use with exchange sex | 16 (37.21) |
| Injection drug use- lifetime | 25 (14.88) |
| Injection drug use- recent | 14 (40.00) |
| Ever had HIV test | 157 (90.75) |
| Last HIV test | |
| Last 3 months | 106 (67.52) |
| 3–6 months ago | 39 (24.84) |
| 6+ months ago | 12 (7.64) |
| Ever had STI test (other than HIV) | 49 (28.32) |
Chi-Square
After descriptive statistics were completed for homeless and foster care demographics, chi-square tests were run to determine if differences existed by type of tie and alter behavior when youth did or did not have alters who were also in foster care. Table 5 demonstrates that youth with home-based peers had more alters without a history of foster care (χ2 = 6.98, df = 1, p = .01). Youth with alters who objected to condom use had more alters without a history of foster care (χ2 = 8.35, df = 1, p = .01). Youth with alters who used heroin (χ2 = 13.08, df = 1, p < .001) or used injection drugs (χ2 = 7.00, df = 1, p = .01) had more alters with a history of foster care.
Table 5.
Egocentric Network: Alter Foster Care Status by Social and Behavior Category (n = 163)
| Foster Care Alter | Non-Foster Care Alter | Chi-Square | |
|---|---|---|---|
| n (%) | n (%) | χ2 | |
| Social Network Engagement | |||
| Type of Tie | |||
| Home-based | 18 (11.46) | 43 (27.39) | 6.98** |
| Partner | 6 (3.92) | 16 (10.46) | 1.16 |
| Friend | 13 (8.33) | 72 (46.15) | 1.34 |
| Staff | 4 (2.61) | 14 (9.15) | 0.09 |
| Talk to about sex | 14 (9.15) | 46 (30.07) | 0.58 |
| Talk to about HIV testing | 14 (9.21) | 27 (24.34) | 2.88 |
| Someone they can confide in | 13 (8.50) | 48 (31.37) | 0.19 |
| Someone they can get advice from | 13 (8.50) | 47 (30.72) | 0.47 |
| Alter Behavior | |||
| Condomless sex | 5 (3.31) | 18 (11.92) | 0.06 |
| Objects to condom use | 9 (5.88) | 12 (7.84) | 8.35** |
| Uses meth | 6 (4.11) | 12 (8.22) | 2.06 |
| Uses heroin | 5 (3.40) | 2 (1.36) | 13.08*** |
| Uses cocaine | 4 (2.67) | 7 (4.67) | 2.21 |
| Uses injection drugs | 4 (2.70) | 3 (2.03) | 7.00** |
p < .05,
p < .01,
p < .001.
Logistic Regressions
For the final multivariable logistic regressions, two-tailed tests were conducted using the existing literature on risks associated with particular foster care experiences. For the final analysis, ten fewer respondents were included in the logistic regressions. Demographic and chi square results largely relied on survey data, while egocentric data was collected separately. Only those participants that answered both surveys were included in the logistic regressions. Controls included in the models in Table 2 were age, race, gender, sexual orientation, placement type, feelings of support while in foster care, and feelings of respect while in foster care. Significant results from the logistic regressions indicated that youth with more time spent in foster care were significantly more likely to have alters with a history of foster care (OR = 4.98, CI = 1.49, 16.62), youth with higher numbers of foster care placements were more likely to have home-based alters (OR = 2.25, CI = 1.03, 4.89), and youth experiencing homelessness for the first time before exiting foster care were less likely to have alters they considered friends (OR = .38, CI = .19, .80). Regarding alter behaviors, youth experiencing homelessness for the first time before exiting foster care were more likely to have alters engaging in condomless sex (OR = 4.65, CI = 1.40, 15.50), and youth with higher numbers of foster care placements were more likely to have alters engaging in methamphetamine use (OR = 3.75, CI = 1.16, 12.19
Discussion
After examining the results of an exploratory analysis of homeless former foster youth, a number of interesting findings emerged. A series of chi-square tests were run to determine if network engagement significantly differed according to the foster care status of the youths’ alters. Results indicated that youths whose alters did not have a history of foster care were more likely to have home-based alters, indicating their connection to individuals from home. These youths also had more alters who objected to condom use, indicating certain risky sex behaviors among the alters without a history of foster care. Youths whose alters did have a history of foster care had more alters engaging in drug use behaviors, including methamphetamines and injection drug use. Previous literature has shown that some foster care experiences are significantly associated with methamphetamine use and having alters who engage in methamphetamine use (Yoshioka-Maxwell et al., 2015). Additionally, research has indicated that the association between foster care experiences and risk outcomes (such as HIV risk) can serve as both protective and risk factors; depending on the context of the situation, some experiences present as risk factors while serving as protective factors in other situations (Yoshioka-Maxwell & Rice, 2019).
Further analyses indicated that foster care experiences such as time spent in foster care, number of foster care placements, and homelessness experiences before transitioning out of foster care were significantly associated with the type of alter reported in the youths’ networks. One aspect we were trying to better understand was the role of alters with a shared history of foster care. Thus, it became important for us to examine from where youth knew their alters, in an attempt to determine the point in which these alters originated within a participant’s foster care and homelessness history. Results indicated that whereas time spent in foster care was associated with having alters from foster care, a higher number of placements was significantly associated with more home-based peers. These results suggest that time spent in foster care and number of foster care placements are two different experiences that impact the type of tie the youth engage with after they transition out of care. One potential cause for these differences may stem from the ability to establish network connections. Having a larger number of foster care alters makes sense if a youth has spent more time in placement, whereas a disruption of social networks may arise if youth are moved from on placement to another. This disruption may force youth to connect with home-based peers, given the turnover in social networks that results from multiple placements. As previously discussed, networks are important for youth during transitional periods (Perry, 2006). Better understanding factors that lead to network connectivity or disruption may be pivotal in addressing network-based interventions aimed at reducing risk. Finally, youth with experiences of homelessness before transitioning out of foster care were less likely to have alters they considered friends, perhaps implying that having a risk factor, such as childhood homelessness, impacts the types of alters a youth has and the number of positive and supportive ties in their lives
Concerning alter behaviors, youths with experiences of homelessness before transitioning out of foster care were more likely to have friends engaging in condomless sex, suggesting that earlier experiences of homelessness additionally serve as a risk factor for alter behaviors. Furthermore, youth with higher numbers of foster care placements were more likely to have alters engaging in risk behaviors, indicating that homeless former foster youth experiencing some risk factors while in foster care may connect with youth who engaged in HIV-risk behaviors. These results suggest that certain foster care experiences and risk factors within foster care negatively impact the types of alters with which youths engage. This result connects back to some of the tenets of RAAM, indicating that negative contact with socializing agents as amplifying risk while positive contact with socializing agents as abating it (Milburn et al., 2009). Finally, time spent in placement was not significantly associated with alter behaviors, potentially indicating that the effect of network disruption, evident from homeless experiences and multiple foster care placements, negatively impacts the types of alters with which youths are connected. Frequent network disruptions may be associated in some way with engagement with risky alters.
Limitations
Several limitations exist for the analyses conducted. First, the data represent a cross-sectional analysis of homeless youth from Los Angeles. This cross-sectional nature indicates that causality cannot be implied. Further, homeless youth in Los Angeles do not necessarily represent the characteristics of youth across the country, and foster care experiences, whereas having many factors in common, vary across counties and states. Thus, foster care experiences may have a considerable amount of variation depending on the location of the placement. Finally, whereas these analyses include a range of experiences common to many former foster youth, with the sample of homeless former foster youth used, understanding which experiences may also be common among homeless youth without a history of foster care is difficult. To truly understand which experiences drive network engagement, a sample of homeless youth with and without a history of foster care would need to be compared. Additional information would also need to be gathered to disentangle the experiences of foster care with the experiences of homelessness, and where the two converge. Without additional data collection efforts and data analysis, it is difficult to determine causality. Additionally, model fit statistics were run for all models, indicating weak pseudo r-square values for these models (all less than .20). While not ideal, given the amount of exploration, rather than confirmation, conducted in this analysis, it was not unexpected. While the model fit statistics may not have been what we had hoped for, the significance of associations is indicative of important associations to be further examined. Finally, whereas all data were gathered through self-report measures and subject to a number of biases, this analysis is more concerned with youths’ perceptions of experiences, rather than information that may be more consistent with their placement records.
Implications
The results of these analyses suggest that specific foster care experiences impact the types of alters homeless former foster youth identify in their social networks and the types of behaviors in which their networks engage. The importance of these interactions lies in the significance of network engagement in overall outcomes for youth. Knowledge gathered from RAAM and previous research (Milburn et al., 2009) suggests that negative contact with a socializing agent amplifies individual risk, whereas positive contact abates this risk. As a result, focus on the network engagement of this population is required to understand the role of network alters in amplifying or abating risk. From the current study, experiences of foster care impacted the types of potentially positive interactions the youths had with socializing agents, such as alters from home or alters they consider friends, while putting youth at risk for negative interactions with socializing agents who engage in risky sex and drug behaviors. These findings can be useful in targeting interventions for reducing risk among this population, perhaps indicating the point of intervention where types of social networks, and their associated risk, may be mitigated. Additionally, because network behaviors impact an individual’s behaviors, these analyses have far-reaching implications for the behavioral health of homeless former foster youth. These results can be broadly applied to a number of health and behavioral health interventions. Thus far, research has focused on the use of social networks to target health outcomes among homeless youth but has not considered those experiences unique to youth with a foster care history. For example, the peer-led HIV-prevention intervention, “Have You Heard?” (Rice, Tulbert, Cederbaum, Barman-Adhikari, & Milburn, 2012), incorporates peer networks to impact HIV-risk behaviors. Initial testing has shown positive effects for HIV testing, but this intervention does not incorporate differences in the social networks of homeless former foster youth or the fact that youth with a history of foster care have high rates of HIV-risk behaviors, including substance use. Without fully understanding the impact of foster care experiences on homeless youth, specifically targeting the risk factors common to this population is difficult. As a field of research, social work would benefit from focusing research efforts on including social network engagement into behavioral health interventions may more holistically address factors impacting behavioral health and aid us in better understanding the impact of life histories on network engagement. Additional considerations within the child welfare system should also be made, to better understand how reduction of some of the risk factors associated with foster care, such as homelessness during foster care, number of placements, and time spent in foster care may positively impact both types of network engagement and engagement in risk behaviors into adulthood. Future policy research is needed to better understand the long-term implications of child welfare policy on young adult homelessness.
Table 6.
Egocentric Logistic Regressions: Foster Care Experiences by Social and Behavior Categories (n = 163)
| Time in Foster Care | Number of Foster Care Placements | Foster Care Exit Age | First Homelessness - Before Foster Care | |
|---|---|---|---|---|
| n = 153 | n = 153 | n = 153 | n = 153 | |
| OR (95%CI) | OR (95%CI) | OR (95%CI) | OR (95%CI) | |
| Social Network Engagement | ||||
| Type of Tie | ||||
| Home-based | 2.13 (.92, 4.62) | 2.25 (1.03, 4.89)* | 1.53 (.28, 8.32) | 1.76 (.84, 3.67) |
| Foster care (any) | 4.98 (1.49, 16.62)** | 1.03 (.38, 2.83) | 1.43 (.15, 13.97) | 1.67 (.63, 4.41) |
| Partner | 1.08 (.32, 3.64) | 3.32 (.90, 12.28) | >999 (<.001, >999) | 1.21 (.36, 4.03) |
| Friend | .61 (.30, 1.27) | 1.44 (.68, 3.05) | .32 (.06, 1.67) | .38 (.19, .80)** |
| Staff | 1.51 (.44, 5.22) | 2.00 (.58, 6.91) | >999 (<.001, >999) | 1.09 (.34, 3.51) |
| Talk to about sex | .56 (.26, 1.19) | .93 (.43, 2.04) | 1.65 (.79, 3.42) | .79 (.38, 1.65) |
| Alter Behavior | ||||
| Condomless sex | .79 (.38, 1,65) | 1.53 (.52, 4.50) | 1.29 (.14, 11.79) | 4.65 (1.40, 15.50)** |
| Objects to condom use | 1.29 (.18, 9.16) | .83 (.10, 6.90) | .08 (.00, 1.92) | 6.11 (.54, 68.78) |
| Uses meth | 1.10 (.37, 3.31) | 3.75 (1.16, 12.19)* | 1.13 (.12, 10.86) | 2.00 (.67, 5.97) |
| Uses heroin | 1.23 (.16, 9.68) | 2.56 (.34, 19.24) | .06 (.00, 1.54) | 3.55 (.34, 36.91) |
| Uses cocaine | .53 (.13, 2.20) | 1.99 (.45, 8.72) | .42 (.04, 4.40) | 4.18 (.79, 22.00) |
| Uses injection drugs | .99 (.15, 6.7) | 1.56 (.23, 10.64) | >999 (<.001, >999) | 4.17 (.41, 42.18) |
p < .05,
p < .01,
p < .001.
Highlights.
connectivity to other foster youth was associated with engagement in drug use.
More placements was significantly associated with more home-based peers.
Early homelessness was associated with connection engagement in condomless sex.
More placements was associated with connections engaging in risk behaviors.
Acknowledgements:
Research reported in this publication was supported by the National Institute of Mental Health of the National Institutes of Health under award number F31MH112251. The content is solely the responsibility of the authors and does not necessarily
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of Interest: Amanda Yoshioka-Maxwell has no conflict to disclose.
Human Subjects/Informed Consent: Signed voluntary informed consent was obtained from each youth, with the caveats that child abuse and suicidal and homicidal intentions would be reported. Informed consent was obtained from youth 18 years and older, and informed assent was obtained from youth 14 to 17 years old. The Institutional Review Board (IRB) at the University of Southern California waived parental consent, as homeless youth younger than 18 years are unaccompanied minors who may not have a parent or adult guardian who could give consent
Contributor Information
Amanda Yoshioka-Maxwell, University of Hawaii at Manoa.
Eric Rice, University of Southern California.
References
- Ahrens KR, Richardson LP, Courtney ME, McCarty C, Simoni J, & Katon W (2010). Laboratory-diagnosed sexually transmitted infections in former foster youth compared with peers. Pediatrics, 126(1), e97–e103. doi: 10.1542/peds.2009-2424 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bao WN, Whitbeck LB, & Hoyt DR (2000). Abuse, support, and depression among homeless and runaway adolescents. Journal of Health and Social Behavior, 41(4), 408–420. [PubMed] [Google Scholar]
- Barth RP (1990). On their own: The experiences of youth after foster care. Child and Adolescent Social Work, 7(5), 419–440. [Google Scholar]
- Benjet C, Borges G, Medina-Mora ME, & Mendez E (2012). Chronic childhood adversity and stages of substance use involvement in adolescents. Drug and Alcohol Dependence, 131(1–2), 85–91. doi: 10.1016/j.grudalcdep.2012.12.002 [DOI] [PubMed] [Google Scholar]
- Berzin SC (2008). Difficulties in the transition to adulthood: Using propensity scoring to understand what makes foster youth vulnerable. Social Service Review, 82(2), 171–196. [Google Scholar]
- Black MM, Oberlander SE, & Lewis T (2009). Sexual intercourse among adolescents maltreated before age 12: A prospective investigation. Pediatrics, 124(3), 941–949. [DOI] [PubMed] [Google Scholar]
- Blakeslee J (2011). Expanding the scope of research with transition-age foster youth: Applications of the social network perspective. Child & Family Social Work, 17(3), 326–336. doi: 10.1111/j.1365-2206.2011.00787.x [DOI] [Google Scholar]
- Bousman CA, Blumberg EJ, Shillington AM, Hovell MF, Ji M, Lehman S, & Clapp J (2005). Predictors of substance use among homeless youth in San Diego. Addictive Behaviors, 30(6), 1100–1110. [DOI] [PubMed] [Google Scholar]
- Braciszewski JM, & Stout RL (2012). Substance use among current and former foster youth: A systematic review. Children and Youth Services Review, 34, 2337–2344. doi: 10.1016/j.childyouth.2012.08.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brechwald WA, & Prinstein MJ (2011). Beyond homophily: A decade of advances in understanding peer influence processes. Journal of Research on Adolescence, 21(1), 166–179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carpenter SC, Clyman RB, Davidson AJ, & Steiner JF (2001). The association of foster care or kinship care with adolescent sexual behavior and first pregnancy. Pediatrics, 108(3), e36. doi: 10.1542/peds.108.3.e46 [DOI] [PubMed] [Google Scholar]
- Cauce AM, Paradise M, Embry L, Morgan C, Lohr Y, Theofelis J, … & Wagner V (1998). Homeless youth in Seattle: Youth characteristics, mental health needs, and intensive case management. Community-based programming for children with serious emotional disturbances: Research and evaluation, 611–632. [Google Scholar]
- Chettiar J, Shannon K, Wood E, Zhang R, & Kerr T (2010). Survival sex work involvement among street-involved youth who use drugs in a Canadian setting. Journal of Public Health, 32(3), 322–327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Collins M (2004). Enhancing services to youths leaving foster care: Analysis of recent legislation and its potential impact. Children and Youth Services Review, 26(11), 1051–1065. [Google Scholar]
- Courtney ME, Dworsky A, Brown A, Cary C, Love K, & Vorhies V (2011). Midwest evaluation of adult functioning of former foster youth: Outcomes at age 26. Chicago, IL: Chapin Hall at the University of Chicago. [Google Scholar]
- Courtney ME, Dworsky A, Ruth G, Keller T, Havlicek J, & Bost N (2005). Midwest evaluation of the adult functioning of former foster youth: outcomes at age 19. Chicago, IL: Chapin Hall Center for Children at the University of Chicago. [Google Scholar]
- Daining C, & DePanfilis D (2007). Resilience of youth in transition from out-of-home care to adulthood. Children and Youth Services Review, 29, 1158–1178. doi: 10.1016/j.childyouth.2007.04.006 [DOI] [Google Scholar]
- De La Haye K, Green HD, Kennedy DP, Zhou A, Golinelli D, Wenzel SL, & Tucker JS (2012). Who is supporting homeless youth? Predictors of support in personal networks. Journal of Research on Adolescence, 22(4), 604–616. doi: 10.1111/j.15327795.2012.00806.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- DiClemente RJ, Crittenden CP, & Rose E (2008). Psychosocial predictors of HIV associated sexual behaviors and the efficacy of preventions interventions in adolescents at-risk for HIV infection: What works and what doesn’t? Psychosomatic Medicine, 70(5), 598–605. [DOI] [PubMed] [Google Scholar]
- Ennett ST, Baily SL, & Federman EB (1990). Social network characteristics associated with risk behaviors among runaway and homeless youth. Journal of Health and Social Behavior, 40(March), 63–78. [PubMed] [Google Scholar]
- Fischer PJ, & Breakey WR (1991). The epidemiology of alcohol, drug, and mental disorders among homeless persons. American Psychologist, 46(11), 1115–1128. doi: 10.1037/0003-066x.46.11.1115 [DOI] [PubMed] [Google Scholar]
- Furman W, & Buhrmeter D (1992). Age and sex differences in perceptions of networks of personal relationships. Child Development, 63, 103–115. [DOI] [PubMed] [Google Scholar]
- Geenen S, & Powers LE (2007). “Tomorrow is another problem”: The experiences of youth in foster care during their transition into adulthood. Children and Youth Services Review, 29(8), 1085–1101. [Google Scholar]
- Ghanem K, Hutton H, Zenilman J, Zimba R, & Erbelding E (2005). Audio computer assisted self-interview and face to face interview modes in assessing response bias among STD clinic patients. Sexually Transmitted Infection, 81(5), 421–425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gleghorn AA, Marx R, Vittinghoff E, & Katz MH (1998). Association between drug use patterns and HIV risks among homeless, runaway, and street youth in Northern California. Drug and Alcohol Dependence, 51, 219–227. [DOI] [PubMed] [Google Scholar]
- Greene JM, Ennett ST, & Ringwalt CL (1997). Substance use among runaway and homeless youth in three national samples. American Journal of Public Health, 87(2), 229–235. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Halkitis PN, Kapadia F, Siconolfi DE, Meoller RW, Figueroa RP, Barton SC, & Blachman-Forshay J (2013). Individual, psychosocial, and social correlates of unprotected anal intercourse in a new generation of young men who have sex with men inNew York City. American Journal of Public Health, 103(5), 889–895. doi: 10.2105/AJPH.2012.300963.epub2013Mar14 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hudson AL, & Nandy K (2012). Comparisons of substance abuse, high-risk sexual behavior, and depressive symptoms among homeless youth with and without a history of foster care placement. Contemporary Nurse, 42(2), 178–186. [DOI] [PubMed] [Google Scholar]
- Jones L (2012). Measuring resiliency and its predictors in recently discharged foster youth. Child and Adolescent Social Work Journal, 29, 515–533. doi: 10.1007/s.10560-012-0275z [DOI] [Google Scholar]
- Jones L (2013). The family and social networks of recently discharged foster youth. Journal of Family Social Work, 16(3), 225–242. doi: 10.1080/10522158.2013.786307 [DOI] [Google Scholar]
- Jones R (2003). Survey data collection using audio computer assisted self-interview. Western Journal of Nursing Research, 25(3), 349–358. [DOI] [PubMed] [Google Scholar]
- Joseph MA, O’Connor TG, Briskman JA, Maughan B, & Scott S (2014). The formation of secure new attachments by children who were maltreated: An observational study of adolescents in foster care. Developmental Psychopathology, 26(1), 67–80. [DOI] [PubMed] [Google Scholar]
- Kerr T, Stoltz J, Marshall BDL, Lai C, Strathdee SA, & Wood E (2009). Childhood trauma and injection drug use among high-risk youth. Journal of Adolescent Health, 45, 300–302. doi: 10.1016/j.jadohealth.2009.03.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Keeshin BR, & Campbell K (2011). Screening homeless youth for histories of abuse: Prevalence, enduring effects, and interest in treatment. Child Abuse & Neglect, 35, 401–407. [DOI] [PubMed] [Google Scholar]
- Kidd SE (2003). Street youth: Coping and interventions. Child and Adolescent Social Work Journal, 20(4), 235–261. [Google Scholar]
- Kipke MD, Montgomery SB, Simon TR, & Iverson EF (1997). Substance abuse disorders among runaway and homeless youth. Substance Use & Misuse, 32(7&8), 969–986. [DOI] [PubMed] [Google Scholar]
- Kral AH, Molner BE, Booth RE, & Waters JK (1997). Prevalence of sexual risk behaviour and substance use among runaway and homeless adolescents in San Francisco, Denver, and New York City. International Journal of STD and AIDS, 8, 109–117. [DOI] [PubMed] [Google Scholar]
- Laumann EO, Marsden PV, Prensky D, Burt RS, & Minor MJ (1983). Applied network analysis.
- Macalino GE, Celentano DD, Latkin C, Strathdee SA, & Vlahov D (2002). Risk behaviors by audio computer-assisted self-interviews among HIV-seropositive and HIV-seronegative injection drug users. AIDS Education and Prevention, 14(5), 367–378. [DOI] [PubMed] [Google Scholar]
- McKinney-Vento Homeless Assistance Act of 1988, PL. 100–628, 101 Stat. 482
- McPherson J, Smith-Lovin L, & Cook JM (2001). Birds of a feather: Homophile in social networks. Annual Review of Sociology, 27, 415–444. [Google Scholar]
- Metzger DS, Koblin B, Turner C, Navaline H, Valenti F, Holte S, … Seage GR 3rd. (2000). Randomized controlled trial of audio computer-assisted self-interviewing: Utility and acceptability in longitudinal studies. American Journal of Epidemiology, 152(2), 99–106. [DOI] [PubMed] [Google Scholar]
- Milburn NG, Rice E, Rotheram-Borus M, Mallett S, Rosenthal D, Batterham P, … Duan N (2009). Adolescents exiting homelessness over two years: The Risk Amplification and Abatement Model. Journal of Research on Adolescence, 19(4), 762–785. doi: 10.1111/j.1532-7795.2009.00610.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Morton MH, Dworsky A, Matjasko JL, Curry SL, Schlueter D, Chavez R, & Farrell AF (2017). Prevalence and correlates of youth homelessness in the United States. Journal of Adolescent Health, 62(1), 14–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nyamathi A, Branson C, Kennedy B, Salem B, Khalilifard F, Marfisee M, … Leake B (2012a). Impact of nursing intervention on decreasing substances among homeless youth. American Journal of Addiction, 21(6), 558–565. doi: 10.1111/j.15210391.2012.00288.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Perry BL (2006). Understanding social network disruption: The case of youth in foster care. Social Problems, 53(3), 371–391. [Google Scholar]
- Reilly T (2003) Transition for care: Status and out of youth who age out of foster care. Child Welfare, 82(6), 727–748. [PubMed] [Google Scholar]
- Rice E (2012). [HIV risks in large social networks of homeless youth.] Unpublished raw data
- Rice E, Milburn NG, & Rotheram-Borus MJ (2007). Pro-social network influences on HIV/AIDS risk behaviors among newly homeless youth in Los Angeles. AIDS Care: Psychological and Socio-medical Aspects of AIDS/HIV, 19(5), 697–704. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rice R, Milburn NG, Rotheram-Borus M, Mallett S, & Rosenthal D (2005). The effects of peer group network properties on drug use among homeless youth. American Behavioral Scientist, 48, 1102–1123. doi: 10.1177/0002764204274194 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rice E, Milburn NG, & Rotheram-Borus MJ (2007). Pro-social network influences on HIV/AIDS risk behaviors among newly homeless youth in Los Angeles. AIDS Care: Psychological and Socio-medical Aspects of AIDS/HIV, 19(5), 697–704. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rice E, & Rhoades H (2013). How should network-based prevention for homeless youth be implemented? Addiction, 108(9), 1625–1626. doi: 10.1111/add.12255 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rice E, Stein JA, & Milburn N (2008). Countervailing social network influences on problem behaviors among homeless youth. Journal of Adolescence, 31(5), 625–639. doi: 10.1016/j.adolescence.2007.10.008 [DOI] [PubMed] [Google Scholar]
- Rice E, Tulbert E, Cederbaum J, Barman Adhikari A, & Milburn NG (2012). Mobilizing homeless youth for HIV prevention: a social network analysis of the acceptability of a face-to-face and online social networking intervention. Health education research, 27(2), 226–236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rice E, & Yoshioka-Maxwell A (2015). Social network analysis as a toolkit for the science of social work. Journal of the Society for Social Work and Research, 6(3), 369–383. [Google Scholar]
- Salomonson-Sautel S, Van Leeuwen JM, Gilroy C, Boyle S, Malberg D, & Hopfer C (2008). Correlates of substance use among homeless youth in eight cities. American Journal on Addictions, 17, 224–234. doi: 10.1080/10550490802019964 [DOI] [PubMed] [Google Scholar]
- Schilling EA, Aseltine RH, & Gore S (2007). Adverse childhood experiences and mental health in young adults: A longitudinal survey. BMC Public Health, 7, 30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Thrane LE, Hoyt DR, Whitbeck LB, & Yoder KA (2006). Impact of family abuse on running away, deviance, and street victimization among homeless rural and urban youth. Child Abuse & Neglect, 30(10), 1117–1128. [DOI] [PubMed] [Google Scholar]
- Turner C, Ku L, Rogers S, Lindberg L, Pleck J, & Sonenstein F (1998). Adolescent sexual behavior, drug use, and violence: increased reporting with computer survey technology. Science, 280(5365), 867–873. [DOI] [PubMed] [Google Scholar]
- U. S. Department of Health and Human Services. (2008). The AFCARS report: Preliminary FY 2006 estimates as of January 2008. Retrieved from http://www.acf.hhs.gov/programs/cb/stats_research/afcars/tar/report14.htm
- Vaughn MG, Ollie MT, McMillen JC, Scott L, & Munson M (2007). Substance use and abuse among older youth in foster care. Addictive Behaviors, 32, 1929–1935. doi: 10.1016/j.addbeh.2006.12.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weis J, & Hawkins DJ (1981). Preventing delinquency. Washington, DC: OJJDP. [Google Scholar]
- Wenzel SL, Hsu HT, Zhou A, & Tucker JS (2012). Are social network correlated of heavy drinking similar among black homeless youth and white homeless youth? Journal of Studies on Alcohol and Drugs, 73(6), 885–889. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Whitbeck L, & Hoyt D (1999). Nowhere to grow: Homeless and runaway adolescents and their families. New York, NY: Aldine de Gruyter. [Google Scholar]
- White CR, O’Brien K, White J, Pecora PJ, & Phillips CM (2008). Alcohol and drug use among alumni of foster care: Decreasing dependency through improvement of foster care experiences. Journal of Behavioral Health Services & Research, 35(4), 419–434. [DOI] [PubMed] [Google Scholar]
- Yoshioka-Maxwell A, & Rice E (2016). Exploring the impact of network characteristics on behavioral health outcomes among homeless former foster youth. International Journal of Public Health, 62(3), 371–378. [DOI] [PubMed] [Google Scholar]
- Yoshioka-Maxwell A, & Rice E (2019). Exploring the relationship between foster care experiences and HIV risk behaviors among a sample of homeless former foster youth. AIDS and Behavior, 23(3), 792–801. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yoshioka-Maxwell A, Rice E, Rhoades H, & Winetrobe H (2015). Methamphetamine use among homeless former foster youth: The mediating role of social networks. Journal of Alcoholism and Drug Dependence, 3(2), pii: 197. [DOI] [PMC free article] [PubMed] [Google Scholar]