Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Apr 1.
Published in final edited form as: J Ethn Subst Abuse. 2018 Sep 5;19(2):174–189. doi: 10.1080/15332640.2018.1489748

A Social Network Analysis of Oxford Houses for Native Americans

Leonard A Jason 1, Ed Stevens 2, Jessica Kassanits 3, Angela Reilly 4, Ted Bobak 5, Mayra Guerrero 6, Nathan J Doogan 7
PMCID: PMC6401329  NIHMSID: NIHMS1515125  PMID: 30183538

Abstract

Social network data were collected among residents of an Oxford House (OH) recovery home, which was located on Suquamish Tribal territory. Data were collected on the social connections of eight male residents (including four Native Americans and four non-Native Americans) using a social network instrument (Jason & Stevens, 2017). A number of social network relationship type were examined including: friendship, trust, and mentorship. Social network data assessed included diameter, reciprocity, the average path length, cohesion, density, transitivity, and centrality. Findings indicated that the OH provided residents a well-integrated network with multiple sources of friendship, trust and mentors. This is of importance as recovery from substance abuse is facilitated when recovering individuals are provided stable and well-functioning networks that foster social support, access to resources, and mentorship.

At the 1965 Swampscott meeting, the field of Community Psychology emerged as a new discipline within psychology (Anderson et al., 1966). In one of the key addresses at this meeting, Glidewell (1966) commented that we needed to shift the attention of psychologists to interconnections that values, motives and feelings shape behavior and adaptation. This type of work had been occurring in sociology (e.g., Homans, 1950, 1961; Blau, 1964), and among theorists in social psychology who saw group contexts as ways to study attribution, social exchange, etc. (e.g. Festinger, 1955; Festinger, Schachter, & Back, 1950). But the field of community psychology tried to go beyond ‘groups’, to study ‘communities’ as a geospatial/geosocial entity. This was also true of early work from the 50’s and 60’s based on social network (“sociometry”) conceptions (Leinhardt, 1977).

Over the past few decades, considerable work has dealt with this social environment, with individuals being exposed to what can be represented by a social network, potentially with multiple relational dimensions (e.g., friend, mentor, adversary). This approach has led to major advances in our understanding of the role of peer affiliations in substance use among adolescents (Brechwald & Prinstein, 2011; Dishion, 2013). Schools provide social scientists with natural social laboratories because of their organization of youth into same-age cohorts, which often represent the majority of youths’ social contexts. For example, Weerman et al. (2011) found the average delinquency level of someone’s friends in the school network had a significant effect on delinquent behavior of the respondents, and leaving or joining informal street-oriented youth groups had a substantial effect on changes in delinquency.

Social networks have also received increasing attention by researchers in the field of substance use disorders. Personal network studies of substance use recovery have established the relevance of participant-reported associates as facilitators of treatment entry (Davey, Latkin, Hua, Tobin, & Strathdee, 2007; Kelly et al., 2010) and mediators of ongoing sobriety (Humphreys & Noke, 1997; Humphreys, Mankowski, Moos, & Finney, 1999; Kaskutas, Bond, & Humphreys, 2002; Longabaugh et al., 1995). For example, Hahm et al. (2012) found an association between social network characteristics and binge drinking from adolescence to young adulthood. In addition, Mercken et al. (2012) found that similarity in smoking behavior among adolescent friends could be caused by selection of friends based on behavioral similarity, or by influence processes, where behavior is changed to be similar to that of friends. The evolving social network methodologies represented in these studies could be used to help researchers measure and explain the dynamic interplay between friendship, trust, mentoring relationships and behavior change. This approach can simultaneously identify the active social ingredients of recovery house living.

Acute substance abuse treatment may include detoxification and some basic counseling, and treatment is typically short (a few weeks or so). Follow-up stays in post-treatment settings with supportive cohesive peer networks encourage personal transformation, and have been shown to substantially reduce relapse rates (Jason, Olson, & Harvey, 2015; Laudet et al., 2009; Schaefer et al., 2011). Sober living homes are currently the largest recovery-specific, community-based support options for post-treatment recovery (Polcin et al., 2010). Traditional recovery homes employ house managers (paid staff with degrees in substance abuse counseling) who run house meetings, enforce rules, make decisions regarding eviction due to rule violation, collect weekly rent, and oversee the overall operation of the houses. Oxford House (OH) comprises a distinct model of sober living homes; there are no professional staff associated with OHs (Jason, Olson, & Foli, 2008). OHs are rented, gender-specific single-family homes for 6 to 12 individuals. Each OH operates democratically with majority rule (i.e., > 80% approval rate) regarding membership and most other policies (Oxford House Manual, 2011). Residents must follow three simple rules, namely, pay rent, abstain from using alcohol and other drugs, and contribute to the maintenance of the home. Violation of the above rules results in eviction from the House (Oxford House Manual, 2011). The OH organization is the only substance abuse recovery home model endorsed as evidenced-based by the Substance Abuse and Mental Health Services Administration (SAMHSA, 2011).

The DePaul University research group has conducted a number of outcome studies of these OH recovery homes (Jason, Davis, Ferrari, & Anderson, 2007; Jason, Olson, Ferrari, & Lo Sasso, 2006). For example, in a national sample of OH residents, only 18.5% of the participants reported any substance use over one year (Jason, Davis, et al., 2007). Additionally, over the course of the study, the proportion of abstainers in individuals’ personal social networks increased. Another study, Jason et al. (2006) successfully recruited 150 individuals who completed treatment at alcohol and drug abuse facilities in the Chicago metropolitan area. Half of the participants were randomly assigned to live in an OH, while the other half received community-based aftercare services. At a two-year follow-up, there was significantly lower substance use for OH (31.6%) than Usual Care participants (64.8%). Furthermore, OH residents were more likely to be employed (76.1% vs. 48.6%) and less likely to report illegal activities (0.9% vs. 1.8%). This study also found that individuals who stayed at least six months had lower relapse rates and significantly better indicators of personal change such as employment, abstinence self-efficacy, and a larger proportion of abstinent others in the personal “significant persons” network (Jason, Olson et al., 2007). Because OHresidents typically attend 12-step groups, it is probably the case that both the OH model and 12-step involvement provide social support for ongoing recovery. Groh, Jason, Ferrari, and Davis (2009) found that among individuals with high 12-step involvement, the addition of OH residence significantly increased the odds of abstinence (88% vs. 53%). Results suggested that the joint effectiveness of these mutual-help programs promotes very high levels of abstinence.There is evidence that house residents who are friends with at least one other resident are less likely to leave OH recovery homes early (Jason, Stevens, et al., 2012). Brereton et al. (2014) found that the presence of recovery home members in personal social networks predicted retention in the recovery home. Jason, Light, Stevens, and Beers (2014) collected baseline and three-month follow-up data on residents from five OH recovery houses, and found that trust within groups tended to develop over time in part as a function of inter-individual exposure. OH recovery homes with a self-governance structure might promote such interdependence (Light, Jason, Stevens, Callahan, & Stone, 2016; Schachter, 1951). Taken together, these findings strongly point to resident social integration in the house system/culture as a major factor in preventing early dropout and in facilitating a sustained recovery.

These types of recovery homes might be effective with Native Americans, who suffer.a disproportionate burden of mental health problems including substance abuse, posttraumatic stress, violence, and suicide (Gone & Trimble, 2012; SAMHSA, 2007). High prevalence rates of alcohol and illicit drug use has been attributed to current social and cultural factors along with past abusive practices including when colonists from Europe made alcohol easily available to Native Americans (Beauvais, 1998). However, patterns of increases and decreases of substance use over time have been similar, suggesting that Native American youth could be part of the larger adolescent culture (Beauvais, Pamela Jumper-Thurman, & Burnside, 2008). Clearly, there are complex reasons for these disparities in substance use, and as an example, Goldstein, Oetting, Edwards, and Garcia-mason (2009) suggest that youth from relatively isolated environments have higher susceptibility to drugs when they enter a nonreservation in urban areas.

A prior study with Native American participants who were not on reservations found there were no significant ethnic differences between this group and Caucasians, African Americans, and Latinos in initial length of stay in OH, length of alcohol or drug sobriety, or substance use outcomes four months after the initial assessment (Kidney, Alvarez, Jason, Ferrari, & Minich, 2011). That study, however, assessed the experiences of Native Americans living in OHs on non-tribal lands. Our current study examined the social networks within the first OH that opened on a tribal reservation. Given the preliminary and descriptive nature of the current investigation, we graphically assessed friendship, trust and mentorship by examining social network data (e.g., diameter, reciprocity, the average path length, cohesion, density, transitivity, and centrality). Our study was exploratory in nature in that we were attempting to describe elements of social networks within a Native American Oxford House.

Method

Suquamish Tribe

The Suquamish Tribe is located close to Seattle, WA. Our efforts to work with the OHs on these tribal grounds are reported elsewhere (Jason et al., 2017). In that study, we found that the Tribal Council as well as the residents perceived these recovery homes as very compatible with the values and norms of their communities, particularly since their governance system was democratic, and everyone in the houses had an opportunity to be involved in decisions. For the current study, we had interviewed eight of the ten men who had lived in this house during the two-month period of data collection.

Sociodemographic Characteristics of Residents

Residents had lived in the OH from 2 weeks to 14 months (M = 7.5 months). Four were tribal members (two of these four were from the Suquamish tribe and two were from other tribes) and four were not tribal members. The residents indicated their drug(s) of choice before entering the OH as the following: three indicated methamphetamine, two indicated alcohol, two indicated methamphetamine/heroine, and one indicated cocaine, methamphetamine, and alcohol. They had been abstinent from one to 48 months (M = 16 months). The men had been convicted of a crime from one to 60 times (M = 13.4). Four had never been married, two were divorced, one was separated, and one was widowed; they had from 0 to four children (M = 1.5). Regarding education, one had a college education, four had some college, and three had a high school degree. They were an average of 39.3 years old (range from 24 to 49). One worked full-time, two worked part-time, two were retired, and three were unemployed (but several of those unemployed did mention that they were involved in seasonal fishing).

Instrument

The Social Network Instrument measures three types of relationships: friendship, mentoring, and trust, which tap theoretically-significant relationships within the house, comprising the house social structure. It has been used in several previous OH investigations (Jason, Light, Stevens, & Beers, 2014; Jason & Stevens, 2017; Light, Jason, Stevens, Callahan, & Stone, 2016). Each item was rated on a 5-point (0–4) scale. Friendship, which taps non-judgmental social support, was determined by asking, “How friendly are you with this person?” Response options included the following: close friend; friend; acquaintance; stranger; adversary. To assess Trust, residents were asked, “If this person asked to borrow money from you, how much would you be willing to lend them?” Response options included $0, $10, $50, $100, and $500. Mentoring which involves going to a person for advice was assessed by the question, “How often do you go to this person for advice on your recovery and other important life issues?” Responses included very often, often, regularly, rarely or never.

The two individuals (participants 9006 and 9007 in Figures 1, 2 and 3) residing in the house but who didn’t participate in the study were also included in a participant’s network. Thus, these individuals have zero out degrees by definition. Out degrees involves a member or ego making a rating about another member or alter. In social network research, an ego is a single individual making ratings, and the alters are the individuals in his or her network that the ego is rating.

Figure 1:

Figure 1:

Open in a new tab

Friendship: Non-judgmental Social Support

Figure 2:

Figure 2:

Open in a new tab

Trust: Providing Tangible Resources

Figure 3:

Figure 3:

Open in a new tab

Mentoring: Seeking Advice from Peers

Social network relationships have been found to be reliable (Hlebec & Ferligoj, 2002). In our work with this instrument, using a larger sample of 229 residents of OH recovery homes, the Cronbach’s alpha was .85 and all items contributed positively (Jason & Stevens, 2017). We also performed multi-level CFA on the social network instrument and found excellent fit and per-item loading contribution. This social network instrument was correlatedto length of stay in a recovery home and quality of life, and neither age nor sex were significantly correlated with this instrument (Jason & Stevens, 2017). In our larger OH studies, we have used this instrument, and we applied the same social network instrument to the current OH on tribal land. For the purposes of this study, the network relationship variables were dichotomized to facilitate the use of a rich collection of network statistics designed for use with dichotomous networks. The present study used a scoring/coding system that was employed in previous investigations (Jason et al., 2014; Jason & Stevens, 2017; Light et al., 2016), and thus it is logical to dichotomize network variables for the sake of being consistent across studies. A friendship relationship was considered present if the respondent identified a peer as a close friend or friend, and not present otherwise (i.e., acquaintance, stranger, or adversary). A trust relationship was considered present if the respondent reported being willing to loan the peer $100 or $500, and was not present otherwise (i.e., $0, $10, or $50). A mentoring relationship was considered present if the respondent reported seeking advice from the peer very often or quite often, and was not present otherwise (i.e., regularly, rarely or never).

Network Properties

Our study involved whole, directed networks, i.e. all types of relationships were observed bi-directionally, from the viewpoint of both members of each dyad (ego, the “rater”, and alter, the “rated”). In the network contexts, we refer to a resident as a vertex (or node). Relations are referred to as edges, i.e. connections between the nodes of a dyad. Each network is visually represented in its entirety (see Figures 1, 2, and 3). The network statistics of interest each characterize the whole network in a distinct way with a single value; the network-level statistics are described below.

When examining the shortest path length from every node to all other nodes, the diameter is the longest of these path lengths. It is representative of the linear size of a network, and it is the maximum number of social steps a piece of information would need to travel to get from one person in the network to any other person. The average path length is the average number of steps along the shortest paths for all possible pairs of network nodes, and it is a measure of the efficiency of information transfer within a network. One is the lowest and is most efficient possible value, while higher values indicate a larger number of steps needed for information to traverse the network.

Two edges are reciprocal if one goes from resident A to resident B, and another from B to A. Our measure of network reciprocity is the proportion of all network edges for which a reciprocal edge is present. Density is the sum of the edges divided by the number of possible edges. The measure is bounded between zero and one. A value of one means every directed relationship is present, and a lower value means fewer interconnections. Reciprocity is related to density, but reciprocity could be high even if density is not, indicating a strong tendency for mutual connection.

Cohesion is the minimal number of vertices in a social network that need to be removed to disconnect the group. This statistic therefore captures how fragile the network is to the departure of vertices (residents) from the network (house). If a single bridge connecting two subgroups within a house leaves the house, then the disconnected and minimally communicating subgroups might be cause for concern regarding house stability.

Transitivity quantifies the prevalence of a particular pattern among triads of network vertices. Specifically, if A directs an edge to B, and B directs an edge to C, then A also directs an edge to C. A high level of transitivity, per se, usually represents a high level of network clustering, or well-connected groups within the network.

Centrality is often used to identify the most influential person(s) in a social network (i.e., a person that is a focal point or main figure in the group of people being considered). The statistic we used quantifies the degree to which communication through the network depends on a set of mediators through which vertices can get information from other vertices. A high value would indicate that vertices tend to interact through a small number of mediators, rather than directly. In a pure democracy, lower numbers would be better, as any two people have fewer intermediaries with control over the information traversal between them.1

Results

Friendship

Figure 1 describes the friendship network. Among the 10 vertices, there were 47 edges (See Table 1 for a summary of these data). In other words, the average vertex was rated highly enough to be considered a friend by 5 other vertices. Even the 2 individuals who did not fill out the network instrument had in-degree friendship ratings and thus were well liked by others who did fill out the instrument. The diameter of the network was 3 (e.g., information from person 9009 had to traverse three edges to reach the most socially distant other). From a friendship point of view, person 9009 is somewhat distant from others in the house. The average number of steps along the shortest paths for all possible pairs of network nodes was 1.21, and this means on average that information that typically flows through a friendship network flows fairly directly from one vertex to another in this network. The network’s reciprocity was .55 and it was affected by the lack of participation by 9006 and 9007. On average, otherwise, the relationships were reciprocated. Cohesion was 0 because one node, 9009, would be completely isolated from the group (with respect to friendship) if 9004 left the house. Density was .52, and it was also affected by non-participation of 9006 and 9007. Thus, the density for the house is about half of its theoretical maximum. Transitivity was .89, which is likely to be more reflective of the high density of that central group in the graph than of hierarchy. Centrality was 1.20, suggesting that the network is structured such that mediators are infrequently required to transmit information from one node to another. Again, the relative disconnectedness of vertex 9009 (except through 9004) may have influenced this statistic upwards.

Table 1.

Network Measures

Friend Trust Mentor
Vertices 10.00 10.00 10.00
Edges 47.00 38.00 16.00
Mean Degree 9.40 7.60 3.20
Diameter 3.00 3.00 2.00
Average Path Length 1.21 1.49 1.43
Reciprocity 0.55 0.32 0.25
Cohesion 0.00 0.00 0.00
Density 0.52 0.42 0.18
Transitivity 0.89 0.80 0.32
Centrality 1.20 3.20 1.20
Open in a new tab

Trust

Figure 2 describes the trust network—a network of hypothetical resource sharing. Among the 10 vertices, there were 38 edges. In other words, on average residents were willing to lend at least $100 to four others in the house. The diameter was 3 (but going through two people is not that meaningful since it does not measure whether or not those other people would lend a resident money). The average number of steps along the shortest paths for all possible pairs of network nodes was 1.49 (but this is also not likely to be a meaningful dimension). The network’s reciprocity was .32, which places it lower than that found for friendship. Cohesion was 0, and this is consistent with at least one person having a single source of loanable funds. Density was .42, and this is consistent with reciprocity with a score lower than friendship, and implies that residents are more restrictive with whom they will lend money than who they will call their friend. Transitivity was .80, which reflects that people are willing to loan to friends of friends. Centrality was 3.20, meaning that trust is more centralized, thus some residents had access to more sources of funds, but all but one had access to loans.

Mentoring

Figure 3 measures mentoring, which involves seeking advice from a peer. Among the 10 vertices, there were 16 edges. On average, a person had slightly less than 2 mentorship relationships. Having about 2 mentorship relationships could mean 3 distinct things: A is mentor to B and C; A has two mentors, B and C; and A is mentor to B, and C is mentor to A. The diameter was 2 (but given that, on average, a mentor relationship is likely to be a direct relationship). The network’s reciprocity was .25, so willingness to confide in someone was not necessarily reciprocated in mentee/mentor relationships. Cohesion was 0, and this indicates at least one individual with only a single mentorship relationship served either as an indegree or outdegree. Density was .18, which is half of what friendship is, and this is understandable given the smaller number of mentorship relationships that one might expect. The average path length was 1.43 which just reflects the idea that people are not mentors of mentors, and as Transitivity was 0.32, this would be expected, as one would generally not look to a mentor’s mentor to be a mentor. Centrality was 1.20, indicating there was no dominant or central figure in this house representing a unique position of mentorship.

Discussion

Our study found a well-integrated social network of OH residents by examining the diameter, the average path length, reciprocity, cohesion, density, transitivity, and centrality of the house network. This recovery home located on a Suquamish Tribe provided its residents with multiple sources of friendship, trust, and mentorship. This is of importance as recovery from substance abuse is facilitated when individuals are provided stable and well-functioning networks (Jason, Olson, & Foli, 2008). This pilot network study demonstrates elements of social networks that are likely instrumental in supporting Native American Oxford House residents. Although residents of this OH followed the three essential rules for an Oxford House, evidence for the longer term efficacy will need the collection of follow-up data. Certainly there is a need for future studies to compare social network elements among Native Americans living in OHs that are (vs. are not) comprised mostly of Native Americans. At the very least, findings in the present investigation provide a strong basis for future research in this area, among a marginalized sub-population of persons with substance use disorders.

In a prior study, Native Americans expressed some disharmony when they were minorities living within their OHs that were located on non-tribal territory (Kidney et al., 2011). However, the OH in the current study had multiple Native Americans residents, and it was culturally sensitive to the tribal community, as cultural symbols and ceremonies (e.g., talking circle, smudging, sweat lodge) were incorporated organically into the customs of the OH (Jason et al., 2017). In this way, the OH residents inherently incorporated cultural traditions in their house as healing elements, as has occurred in other settings (Gone, 2011, 2013a, b). Acceptance by the tribal community was very high, and these perceptions by the surrounding neighborhood probably positively affected the outcomes reported in this network study. Other interventions have also successfully incorporate this group’s culture such as medicine wheels, sweat lodges and talking circles (Jones-Saumty, Thomas, Phillips, Tivis, & Nixon, 2003; Naquin, Trojan, O’Neil, & Manson, 2006).

Using a confirmatory factor analysis, Jason and Stevens (2017) had previously found that friendship, trust, and mentorship had substantial load factors (.84, .79, and .81) with an indicator of social networks. In the current study, these three network indices were examined in more detail as represented in Figures 1–3 and described in Table 1. For example, nodes were named on average about five times as a friend; four times for trust, but slightly less than two times for a mentorship relationship. This suggests that friendship and trust are more common than a mentorship relationship. In addition, friend relationships were reciprocated, with a score of .55, but reciprocity was only .25 for mentorship, so being willing to mentor someone was not necessarily reciprocated in mentee/mentor relationships. For trust, the network’s reciprocity was .32, which places it between friendship and being a mentor in terms of reciprocity.

These findings are compatible with the findings of Jason, Light, Stevens, and Beers (2014), who found friendships in OHs tend to symmetrical whereas mentorship relationships tend not to be reciprocated. In another investigation, Light, Jason, Stevens, Callahan, and Stone (2106) included three levels of trust (low trust was defined as being willing to lend less than $50, medium trust as $50–$100, and high trust as $100 or more). Light et al examined three outcomes: medium trust, high trust, and mentorship (which were termed confident) relationships. Mentorship relationships were indeed more likely if ego had high trust for the alter, which was consistent with the previous results. The authors concluded that medium trust could serve as a “threshold” which, once achieved in a relationship, is more likely to increase. In other words, the formation of medium trust relationships are particularly important to house social integration processes, and this was also found in the current study.

For friend relationships, density was .52, whereas for mentorship, density was .22, or about half of what was found for friendship, and this is understandable given the smaller number of mentorship relationships that one might expect. For trust, density was .42, and this is consistent with reciprocity being between friendship and mentoring. For friendships, transitivity was .89, which reflects the high density of that central group in the figure, and it was .80 for trust, which reflects that nodes are willing to loan to friends of friends. But for mentorship, transitivity was .32, and this would be expected, as one would generally not look to a mentor’s friend to be a mentor. A highly embedded individual has strong friendship, trust, and role model relationships with other house residents, who provide resources and support recovery. Conversely, an individual isolated from the residential social network perceives few recovery resources and may be prone to drop out.

For friendships, mentoring and trust, cohesion was 0, and while individuals are at least connected to one other person, it means that these small networks are fragile, and the exit of one individual could result in an isolate. In addition, for friend and mentorship, centrality was 1.20, meaning that there was no single influential gatekeeper in the social network; however, centrality was 3.20 for trust, as there were more people who were unable to lend, so willing to lend was a bit more centralized, but still most everyone did have some access to loans. In a democratic house like OH, lower numbers are better, as any two people have fewer barriers for interactions involving friendship and mentorship. In other words, for members of this OH, there was no single source of friendship or mentorship

These network measures capture the construct of social embeddedness, which are the web of relationships within particular sociocultural contexts (Granovetter, 1985; Polanyi, 1944). Specifically, it describes the multiple types of social linkages among residents--friendships, trust, and mentoring.. Social bonds emerge from interdependent activities, which may then lead to increasing levels of trust and self-disclosure (VanLear, 1987). In general, this social embedding was likely enhanced by the democratic nature of the OH. In these types of settings, residents have control of their how the house is operated, and they decide who is allowed to be accepted into the house as well as how the house is managed. Findings from a number of the social network variables support this thesis, and help us better understand the types of supportive social relationships found in this OH, which may protect people in recovery from relapse and improve overall substance abuse recovery rates (Beattie & Longabaugh, 1999; Groh et al., 2007; Moos, 2007; Vaillant, 1983). However, it should be noted that one resident, 2008, had out degrees for the three network variables, but no in degrees, and this might have been due to the fact that he had only lived in the OH for two weeks, and it might have taken him more time for others in the house to develop connections with him.There are a number of limitations in the current study, and the fact that only residents of one OH does limit the generalizability of the findings. We also note as a limitation that only one tribe of Native Americans was examined, and that future investigations with diverse Native Americans would help us understand if social network elements might be related to homogenous versus heterogeneous Native American OHs. The present study did not control for income level, and that this might have had some bearing on the measure of “trust.” It’s quite possible that one’s income level would influence the amount of money one is willing to lend another. In addition, we were only able to sample eight of the ten residents, and although the other two non-participating residents did indicate a willingness to be part of this study, we were never able to find a time to complete the interview due to their busy work schedule and possible reluctance to be part of this study. The longer-term outcomes of the residents are unclear, and assessing the residents over time would provide a better understanding of the dynamic relationships between the network variables.In summary, the results of this investigation seem consistent with other OH studies that examine social support and/or social networks (Jason et al., 2014; Jason & Stevens, 2017; Light et al., 2016) There is a need for more research on effective recovery options for this ethnic group, and the present study presents some promsing descriptive social network data regarding the first OH established on Native American tribal lands.

Acknowledgments

The authors appreciate the help of George Duncan, Marty Selvidge, Sammy Mabe, the OH members and tribal officials for their support and help in with this study.

The authors appreciate the financial support from the National Institute on Alcohol Abuse and Alcoholism (grant number AA022763).

Footnotes

1

Our measure could be called betweenness centrality, but for abbreviation we will call it centrality. There are several measures of centrality, which tend to characterize individual nodes, whereas “centralization” is usually used to refer to a characterization of the whole network. Our measure is one of centralization.

Contributor Information

Leonard A. Jason, DePaul University

Ed Stevens, DePaul University.

Jessica Kassanits, DePaul University.

Angela Reilly, DePaul University.

Ted Bobak, DePaul University.

Mayra Guerrero, DePaul University.

Nathan J. Doogan, Ohio State University

References

  1. Alvarez J, Adebanjo AM, Davidson MK,Jason LA, & Davis MI (2006). Oxford House: Deaf affirmative support for substance abuse recovery. American Annals of the Deaf, 151, 418–422. [DOI] [PubMed] [Google Scholar]
  2. Anderson LS Cooper S, Hassol L Klein DL, Rosenblum G & Bennett CC (1966). (Eds). A report of the Boston Conference on the education of psychologists for community mental health. Boston: Boston University Press. [Google Scholar]
  3. Beasley C, Callahan S, Stecker E, Dekhtyar M, Yang C, Ponziano F,… & Jason LA (2017). A qualitative study of transgender women and cisgender men living together in two recovery homes. Manuscript submitted for publication. [Google Scholar]
  4. Beattie MC, & Longabaugh R (1999). General and alcohol-specific social support following treatment. Addictive Behaviors, 24(5), 593–606. [DOI] [PubMed] [Google Scholar]
  5. Beauvais F (1998). American Indians and Alcohol. Alcohol Health & Research World. 22, 253–259. [PMC free article] [PubMed] [Google Scholar]
  6. Beauvais F, Jumper-Thurman P, & Burnside M (2008). The changing patterns of drug use among American Indian students over the past thirty years. The Journal of the NationaCenter, 15, 15–24. [DOI] [PubMed] [Google Scholar]
  7. Blau PM (1964). Exchange & Power in Social Life. Transaction.
  8. Brechwald WA & Prinstein MJ (2011). Beyond homophily: A decade of advances in understanding peer influence processes. Journal of Research on Adolescence, 21, 166–179. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Brereton KL, Alvarez J, Jason LA, Stevens EB, Dyson VB,… & Ferrari JR (2014). Reciprocal responsibility and social support among women in substance use recovery. International Journal of Self-Help & Self-Care, 8, 239–257. PMCID: PMC4269347 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Davey MA, Latkin CA, Hua W, Tobin KE, & Strathdee S (2007). Individual and social network factors that predict entry to drug treatment. American Journal on Addictions, 16, 38–45. [DOI] [PubMed] [Google Scholar]
  11. Dishion TJ (2013). Stochastic Agent-Based Modeling of influence and selection in adolescence: Current status and future directions in understanding the dynamics of peer contagion. Journal of Research on Adolescence, 23, 596–603. [Google Scholar]
  12. Festinger L (1955). Social psychology and group processes. Annual Review of Psychology, 6, 187–216. [DOI] [PubMed] [Google Scholar]
  13. Festinger K, Schachter S, & Back K (1950). Social Pressures in Informal Groups. New York: Harpers. [Google Scholar]
  14. Glidewell J (1966). Perspectives in community mental health. In Anderson LS, Cooper S, Hassol L, Klein DC, Rosenblum G, & Bennett CC (Eds). A report of the Boston Conference on the education of psychologists for community mental health (p. 33–49). Boston: Boston University Press. [Google Scholar]
  15. Goldstein GS, Oetting ER, Edwards R, & Garcia-mason V (2009).Drug use among Native American young adults. International Journal of the Addictions, 14, 855–860 [DOI] [PubMed] [Google Scholar]
  16. Gone JP (2011). The red road to wellness: Cultural reclamation in a Native First Nations community treatment center. American Journal of Community Psychology, 47, 187–202. [DOI] [PubMed] [Google Scholar]
  17. Gone JP (2013a). A community-based treatment for Native American historical trauma: Prospects for evidence-based practice. Spirituality in Clinical Practice, 1(S), 78–94. [DOI] [PubMed] [Google Scholar]
  18. Gone JP (2013b). Redressing First Nations historical trauma: Theorizing mechanisms for indigenous culture as mental health treatment. Transcultural Psychiatry, 50, 683–706. [DOI] [PubMed] [Google Scholar]
  19. Gone GP, & Trimble JE (2012) American Indian and Alaska Native mental health: Diverse perspectives on enduring disparities. Annual review of clinical psychology, 8, 131–160. [DOI] [PubMed] [Google Scholar]
  20. Granovetter M (1985). Economic action and social structure: The Problem of embeddedness.American Journal of Sociology, 91, 481–510. [Google Scholar]
  21. Groh DR, Jason LA, Ferrari JR, & Davis MI (2009). Oxford House and Alcoholics Anonymous: The impact of two mutual-help models on abstinence. In Jason LA, & Ferrari JR (Eds.). Recovery from addiction in communal living settings: The Oxford House model [Special Issue]. Journal of Groups in Addiction & Recovery, 4, 23–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Groh DR, Olson BD, Jason LA, Davis MI, & Ferrari JR (2007). A factor analysis of the Important People Inventory. Alcohol and Alcoholism, 42(4), 347–353. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hahm HC, Kolaczyk E, et al. (2012). Binge drinking trajectories from adolescence to young adulthood: The effects of peer social network. Substance Use & Misuse, 47, 745–756. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hlebec V, & Ferligoj A (2002). Reliability of social network measurement instruments. Field methods, 14, 288–306. [Google Scholar]
  25. Homans G (1950). The Human Group. New York: Harcourt, Brace, & World. [Google Scholar]
  26. Homans G (1961). Social Behavior: Its Elementary Forms. New York: Harcourt Brace Jovanovich. [Google Scholar]
  27. Humphreys K, Mankowski ES, Moos RH, & Finney JW (1999). Do enhanced friendship networks and active coping mediate the effect of self-help groups on substance abuse? Annals of Behavioral Medicine, 21, 54–60. [DOI] [PubMed] [Google Scholar]
  28. Humphreys K, & Noke J (1997). The influence of post-treatment mutual help group participation on the friendship networks of substance abuse patients. American Journal of Community Psychology 25, 1–16. [DOI] [PubMed] [Google Scholar]
  29. Jason LA, Davis MI, Ferrari JR, & Anderson E (2007). The need for substance abuse after-care: Longitudinal analysis of Oxford House. Addictive Behaviors, 32, 803–818. [DOI] [PubMed] [Google Scholar]
  30. Jason LA, DiGangi JA, Alvarez J, Contreras R, Lopez R, Gallardo S, & Flores S (2013). Evaluating a bilingual voluntary community-based healthcare organization. Journal of Ethnicity in Substance Abuse, 12, 321–338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Jason LA, Kassanits J, Reilly A, Bobak T, Guerrero M, Stevens E, Light J, & Doogan N (2017). Recovery homes for Native Americans. Manuscript submitted for publication. [DOI] [PMC free article] [PubMed]
  32. Jason LA, Light JM, Stevens E, & Beers K (2014). Dynamic social networks in Oxford House recovery homes. American Journal of Community Psychology, 53, 324–334. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Jason LA, Olson BD, Ferrari JR, & Lo Sasso AT (2006). Communal housing settings enhance substance abuse recovery. American Journal of Public Health, 96, 1727–1729. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Jason LA, Olson BD, Ferrari JR, Majer JM, Alvarez J, & Stout J (2007). An examination of main and interactive effects of substance abuse recovery housing on multiple indicators of adjustment. Addiction, 102, 1114–1121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Jason LA, Olson BD, & Foli K (2008). Rescued lives: The Oxford House approach to substance abuse. New York, NY: Routledge. [Google Scholar]
  36. Jason LA, Olson BD, & Harvey R (2015). Evaluating alternative aftercare models for ex-offenders. Journal of Drug Issues, 45, 53–68. PMCID:PMC4307799 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Jason LA, & Stevens E (2017). The reliability and reciprocity of a social network measure. Alcoholism Treatment Quarterly, 35, 317–327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Jason LA, Stevens E, Ferrari JR, Thompson E, & Legler R (2012). Social networks among residents in recovery homes. Advances in Psychology Study, 1, 4–12. [PMC free article] [PubMed] [Google Scholar]
  39. Jones-Saumty D (2002). Substance abuse treatment for Native Americans In Ma GX & Henderson G (Eds.) Ethnicity and Substance Abuse (pp. 270–283). Springfield, IL: Charles C. Thomas. [Google Scholar]
  40. Kaskutas LA, Bond J, & Humphreys K (2002). Social networks as mediators of the effect of Alcoholics Anonymous. Addiction, 97, 891–900. 10.1046/j.1360-0443.2002.00118.x [DOI] [PubMed] [Google Scholar]
  41. Kelly JF, Stout RL, Magill M, Tonigan JS, & Pagano ME (2010). Mechanisms of behavior change in Alcoholics Anonymous: Does AA lead to better alcohol use outcomes by reducing depression symptoms? Addiction, 105, 626–636. doi: 10.1111/j.1360-0443.2009.02820.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Kidney CA, Alvarez J, Jason LA, Ferrari JR, & Minich L (2011). Residents of mutual help recovery homes, characteristics and outcomes: Comparison of four US ethnic subgroups. Drugs: Education, Prevention & Policy, 18, 32–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Laudet A, Becker J, & White W (2009). Don’t wanna go through that madness no more: Quality of life satisfaction as predictor of sustained substance use remission. Substance Use and Misuse, 44, 227–252. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Leinhardt S (1977). Social Networks: An emerging paradigm. New York: Academic. [Google Scholar]
  45. Light JM, Jason LA, Stevens EB, Callahan S, & Stone A (2016). A mathematical framework for the complex system approach to group dynamics: The case of recovery house social integration. Group Dynamics: Theory, Research and Practice, 20(1), 51–64. PMCID: PMC4821464 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Longabaugh R, Wirtz PW, Beattie MC, Noel N, & Stout R (1995). Matching treatment focus to patient social investment and support: 18-month follow-up results. Journal of Consulting and Clinical Psychology, 63, 296–307. [DOI] [PubMed] [Google Scholar]
  47. Mercken L, Steglich C, Sinclair P, Holliday JC, & Moore L (2012). A longitudinal social network analysis of peer influence, peer selection, and smoking behavior among adolescents in British schools. Health Psychology, 31(4): 450–459. [DOI] [PubMed] [Google Scholar]
  48. Moos RH (2007). Theory-based active ingredients of effective treatments for substance use disorders. Drug and Alcohol Dependence, 88, 109–121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Polanyi K (1944). The Great Transformation: The political and economic origins of our time. Boston, MA: Beacon Press. [Google Scholar]
  50. Naquin V, Trojan J, O’Neil G, & Manson SM (2006). The therapeutic village of care: An Alaska Native alcohol treatment model. International Journal of Therapeutic Communities, 27, 105–121. [Google Scholar]
  51. Oxford House Inc. (2011). Oxford House Manual: An Idea Based on a Sound System for Recovering Alcoholics and Drug Addicts to Help Themselves Silver Spring, MA: Oxford House World Services, Inc. [Google Scholar]
  52. Polanyi K (1944). The Great Transformation: The political and economic origins of our time. Boston, MA: Beacon Press. [Google Scholar]
  53. Polcin DL, Korcha R, Bond J, Galloway G, & Lapp W (2010). Recovery from addiction in two types of sober living houses: 12-month outcomes. Addiction Research and Theory, 18, 442–455. [Google Scholar]
  54. Schaefer JA, Cronkite RC, & Hu KU (2011). Differential relationships between continuity of care practices, engagement in continuing care, and abstinence among subgroups of patients with substance use and psychiatric disorders. Journal of Studies on Alcohol and Drugs, 72, 611–621. [DOI] [PubMed] [Google Scholar]
  55. Schachter S (1951). Deviation, rejection, and communication. Journal of Abnormal and Social Psychology, 46, 190–207. [DOI] [PubMed] [Google Scholar]
  56. Substance Abuse and Mental Health Services Administration ((2007). Substance use and substance use disorders among American Indians and Alaska Natives. Retrieved from http://www.oas.samhsa.gov [PubMed]
  57. Substance Abuse and Mental Health Services Administration’s National Survey on Drug Use and Health. (2011). Retrieved from: https://www.samhsa.gov/data/sites/default/files/Revised2k11NSDUHSummNatFindings/Revised2k11NSDUHSummNatFindings/NSDUHresults2011.htm#Ch7
  58. Substance Abuse and Mental Health Services Administration’s National Registry of Evidence-Based Programs and Practices (2011). Oxford House. Retrieved from: http://www.nrepp.samhsa.gov/ViewIntervention.aspx?id=223).
  59. Vaillant GE (1983). The Natural History of Alcoholism. Cambridge, MA: Harvard University Press. [Google Scholar]
  60. VanLear CA (1987). The formation of social relationships A longitudinal study of social penetration. Human Communication Research, 13, 299–322. doi: 10.1111/j.1468-2958. [DOI] [Google Scholar]
  61. Weerman FM (2011). Delinquent peers in context: A longitudinal network analysis of selection and influence effects. Criminology: An Interdisciplinary Journal, 49, 253–286. [Google Scholar]

RESOURCES