Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jan 1.
Published in final edited form as: Alcohol Treat Q. 2021 Nov 27;40(2):191–204. doi: 10.1080/07347324.2021.1987179

Individual and Contextual Protective and Risk Characteristics for Residents of Recovery Homes

Leonard A Jason 1, Ted Bobak 2, Mohammed Islam 3, Mayra Guerrero 4, John M Light 5, Mike Stoolmiller 6
PMCID: PMC9075153  NIHMSID: NIHMS1746031  PMID: 35528863

Abstract

Some recovery homes have facilitating relationships and organizational characteristics, and there are also social capital differences among residents of these recovery homes. It is important to better understand the impact of protective and risk individual and house factors on recovery issues among residents of these community-based settings. Individuals from 42 recovery homes were followed for up to six data collection periods over two years. House level latent class analyses tapped relationship and organizational domains and individual level latent class analyses were from derived from elements of recovery capital. Houses that manifested protective factors provided most residents positive outcomes, except those with elevated self-esteem. Houses that were less facilitating had more negative exits, except for those residents who were the highest functioning. Both individual and house characteristics are of importance in helping to understand risk factors associated with eviction outcomes for residents in recovery homes.

Keywords: Social Capital, Latent Class Analysis, House-level, Individual-level, Recovery Homes

The field of Community Psychology has used Ecological Theory to explore human behavior in interaction with the social and cultural contexts (Kelly, 2006). For example, in Kelly’s (1979) study of two schools, boys with a preference for exploratory behaviors had more positive scores on measures of adaptation, but more exploratory behavior was facilitated in the school that had more turnover, where more students entered and left each year. Other fields such as social psychology have also addressed issues of context, such as with Latané’s (1981) Social Impact Theory. These theories in different fields of study focus on person-environment interactions, and that adaption often varies by differences in individuals and settings.

Recovery homes are social environments that are dependent on interactions operating between and among the residents. These settings provide housing to over 250,000 individuals annually (Jason, Wiedbusch, Bobak, & Taullahu, 2020), and they have been found to promote recovery. For example, and in one study of sober living homes, 51% of the sample reported being abstinent from substances at a six month follow up (Polcin & Henderson, 2008). But it is still unclear the reason for differential outcomes of those who live in recovery homes. Exploring the congruence between individuals and the settings in which they belong (Beasley & Jason, 2015), might help investigators better understand why some are helped and others are not from this residential experience.

Social Capital Theory has been used to better understand how social relationships are resources that can help in the recovery process (Best & Laudet, 2010; Cloud & Granfield, 2008; SAMHSA, 2012). Recovery homes may increase social capital by providing friendship bonds to residents (Best & Laudet, 2010), which could then lead to improvements in resident recovery related beliefs and behaviors (e.g., abstinence related self-efficacy, coping, self-esteem, hope). Social network research by Kelly, Stout, Greene, and Slaymaker (2014), Patterson et al. (2020), and Tracy et al. (2012) provide demonstrations of this theory for those with substance use disorders. But different houses might provide different recovery resources for their members, and individuals might have different capacities to make use of these available resources.

A recent social capital study recruited a relatively large sample of 42 recovery homes from three different geographical regions of the US, and the (Jason, Stoolmiller, & Light, 2021) found that even within the recovery home network, there was a significant amount of variation in individual characteristics. Latent profile analysis of eight recovery indicators (i.e., wages, self-efficacy, stress, self-esteem, quality of life, hope, sense of community, and social support) was able to differentiate the members into five latent classes. Class 1 had the highest scores on almost all indicators, and thus was higher functioning than the other classes. Most of the other latent class profiles were roughly parallel and differed only in level. In contrast, Class 3 was at a medium level of social capital indicators, but had an unusually high level of self-esteem, even higher than class 1. Elevated levels of self-esteem with only average social capital indicators were found to increase recovery risk for this group of residents. This study suggests that there might be different recovery characteristics of residents in these homes, and they might have different types of adaptation to different house typologies.

Using the same data set, Jason, Stoolmiller, Light, and Bobak (2021) recently conducted another latent class analyses based on house level data that included average network density on friendship, advice seeking and willingness to loan, as well as house president ratings on house involvement in chapter activities, house savings, house new resident selectivity, and the ability of the typical house resident to pay their monthly rent. One latent class representing 45% of the recovery houses had the highest density of members willing to loan, able to pay their rent, active involvement in outside chapter activities, and this group of houses had the best outcomes including the lowest eviction rate.

These studies suggest that there are meaningful differences in types of residents and types of recovery homes, and that these typologies probably influence recovery outcomes. The current exploratory study attempted to identify house and individual types through which social environments affect health outcomes. We hypothesized that those with riskier individual characteristics would do better in the optimal house type, whereas only those from the highest social capital class residents would be able to function well in the less optimal recovery homes.

Method

Participants

The current study was set in Oxford House (OH) recovery homes, a network of over 3,000 self-governed recovery homes in the US, housing over 20,000 individuals. There are no professional staff in these houses, which are rented, gender-segregated, and housing from six to 12 individuals in recovery. Residents are required to follow three rules: paying their fair share of the rent (usually from $100 to $125 per week), contributing to the maintenance of the home, and abstaining from substance use.

Data were collected from OHs located in North Carolina, Texas, and Oregon. Member-elected house presidents were asked to introduce the study to residents by reading a description of it from a project-provided script; houses were accepted into the study if the house president and all or all but one member agreed to participate. The first thirteen consenting houses from each state were accepted, and three more houses were added for a total of 42.1

Participants were part of a longitudinal study that collected information every four months over two years. Participants were recruited and interviewed by field research staff in face-to-face meetings. Participants were compensated $20 for completing each assessment. Permission to do this study was obtained by the DePaul University Institutional Review Board.

Measures

Justification of the use of these measures as indicators of social capital is provided elsewhere (Jason, Guerrero, Salomon-Amend, Stevens, Light, & Stoolmiller, 2021).

Wages for the last 30 days was computed by taking the square root of wages to reduce positive skewness and it was used as a continuous variable.

World Health Organization Quality of Life Assessment-Brief (Quality of Life; WHOQOL Group, 1998) is a 24-item questionnaire that assesses quality of life across four dimensions: social relationships, environment, physical, and psychosocial. This scale has been validated in substance using populations (Garcia-Rea & LePage, 2010). The subscales varied in their reliability (αs = .89 for social relationships, .84 for environment, .83 for physical, and .83 for psychological). The alpha for the whole measure for our sample was .89.

The Drug Taking Confidence Questionnaire (ASE; Sklar & Turner, 1999) is an 8-item survey that measures self-efficacy in terms of abstinence. Participants are asked to consider eight theoretical high-risk situations and rate how confident they would be of resisting the urge to use a substance given the hypothetical circumstances. This measure had good reliability in our sample (α =.95).

The Rosenberg’s Self-Esteem Scale (SES; Rosenberg, 1965) was utilized to measure the participant’s positive and negative feelings about the self. SES is a widely used 10-item, global self-esteem scale measured on a 4-point Likert Scale ranging from “strongly agree” to “strongly disagree.” Items include “I think I have a number of good qualities,” “I take a positive attitude towards myself,” and “I feel I do not have much to be proud of.” The internal reliability of the SES scale was .92 in our sample.

The Perceived Stress Scale (PSS; Cohen et al., 1983) was utilized to measure the degree to which situations in participants’ lives are appraised as stressful. PSS consists of 4-items measured on a 5-point Likert scale ranging from “never” to “very often.” Items include “how often have you felt that you were unable to control the important things in your life?” and “how often have you felt difficulties were piling up so high that you could not overcome them?” In our sample, the internal reliability of the PSS scale was .73.

The Interpersonal Support Evaluation List (ISEL; Cohen & Wills, 1985) measured three types of actual or perceived social support (tangible, appraisal, belonging). Tangible support refers to instrumental aid which refers to monetary assistance; appraisal support refers to having someone to talk to about one’s problems; and belonging support refers to the availability of people one can do activities with. The ISEL consists of 12-items measured on a 4-point Likert scale ranging from definitely false to definitely true. The internal reliability of the ISEL scale was .88 in our sample.

The Psychological Sense of Community (SOC; Jason, Stevens & Ram, 2015) is a 9-item scale utilized to measure participant’s sense of community. Items include “This Oxford House is important to me” and “For me, this Oxford House is a good fit.” The three subscales are Entity, Membership, and Self, and for our sample, they have Cronbach alphas of .67, .92, and .91, respectively. The SOC scale was used as a whole measure (α = .91).

Snyder’s State Hope Scale (Hope; Snyder et al., 1996) was utilized to measure participants’ current state of hope. The Hope measure contains two sub-scales Agency (α = .94) and Pathways (α = .81). We included a 3-item subscale of hope that measures Environmental Context (Stevens et al., 2014) (α = .97). This 9-item hope scale was analyzed as a whole measure, and for our sample, the α = .90.

Resident Level Demographics at initial survey included age, race/ethnicity [White (reference category) vs. Black, Hispanic, and Other], educational attainment (HS grad or less, some college, and college grad or higher), employment status (unemployed, part time and full time), and time in residence prior to first survey participation, log transformed to reduce skewness and prevent potential problems due to high leverage outliers.

OH Level Characteristics included gender (OH’s are gendered, male and female). House president ratings at wave 1 also assessed 1) house involvement in chapter activities, 2) house savings, 3) house new resident selectivity, and 4) typical house resident financial status. Level of participation in chapter level activities by the following scale: Not at all involved, somewhat involved, much involved, a great deal involved. House savings was assessed by asking the amount of house savings using the following scale: significant debt (>$5,000);some debt ($1–5,000); neither debt or savings; some savings ($1–5,000); significant savings(>$5,000). New house resident selectivity was assessed by assessing the new resident acceptance rate using the following question: Out of every 10 applicants, how many does your house accept? Typical house resident financial status was measured by assessing the poverty level of typical resident population served with the following question: “on average, how many residents have difficulty meeting their monthly financial obligations? These questions on OH characteristics were obtained at the very first wave of data collection, typically as rated by the house president.

We also included the Addiction Severity Index-Lite (ASI-Lite; McLellan et al., 1997) to assess evictions for problems in drug and alcohol use over the past 30 days. The ASI-Lite has been demonstrated to have good validity and reliability (Cacciola, et al., 2007). This instrument along with other information from the house officers was used to create the following exit status variables: positive outcomes (46% left on good terms), negative outcomes (31% either relapsed, or left for disruptive behavior or financial reasons), still a resident (22%), or left with no reason available (1%). In this study, the 0/1 indicator variable created from this information was left for an eviction due to negative outcomes.

Relationship types

The definition of a tie for each relationship has been operationalized and defined previously (Jason, Guerrero, Lynch, Stevens, Salomon-Amend, & Light, 2020). In brief, a friendship tie means the resident doing the rating (hereafter termed “ego”) considers a rated housemate (hereafter termed “alter”) as a friend or close friend, a loan tie means ego is willing to loan $100-$500 to alter and an advice seeking tie means that ego goes to alter very often or often for advice. The density within a house is the proportion of all ties that exist divided by all ties that could exist. The densities are proportions that range from 0–1 and indicate more or less ties of each type within a house. High friendship density means that most people within a house are friends or close friends. High loan density means that most people within a house would be willing to lend $100-$500 dollars to a housemate. High advice seeking density means that most people within a house were very often or quite often go to a housemate for advice.

Procedure and Sample

The derivation of the analytic sample for these analyses has been described previously in connection with confirmatory factor analysis of the recovery indicators and survival analyses of time to exit (Jason, Guerrero, Salomon-Amend, Stevens, Light, & Stoolmiller, 2021). Because of the greater analytic complexity of latent class analysis compared to the previously published confirmatory factor analysis, we decided to eliminate the “time to” element of survival analysis of exits and focus on whether an exit eventually occurred or not. However, to maintain as much comparability as possible with our previous work, we use the same survival analysis sample.

Residents of participating homes were able to enter the study at any point during two years, and data were collected every four months over two years. Participation was determined based on several criteria related to the process of censoring (a feature of survival models). First, participants had to have provided at least one wave of follow up data beyond a first baseline interview; otherwise, their reason for leaving could not be determined and they could not have contributed any information to the exit modeling. Ultimately, our exit analysis used the resident latent classes as predictors of resident probability of an exit.

For each wave of data collection, participants were classified as either evicted from the OH for any reason, or left voluntarily, or still in residence. With six survey waves, a maximum of five follow up time intervals were available for observing and predicting potential exits. Of course, participants entering OHs after wave 1 had fewer follow up intervals available than those in the study at wave 1, but the analysis included all such participants having at least one follow-up interval—that is, at least one period after their first wave of study participation for which we could determine definitively whether they had exited from the house or not and the reason for the exit.

There were 714 residents of the OHs during this period of time, of which 666 (93%) agreed to participate in this investigation. We excluded 64 of these participants from the current study.2 All the LPA models used the same set of 602 individuals (out of 666 total participants, or 90%, henceforth referred to as the initial risk set). The amount of missing indicator data was minimal (97% of participants had complete data) but in line with current standard practice of including partial data, the 21 participants with partial data on the indicators were still included in all analyses.

Although the total sample included 602 participants, only 563 participants had complete data on all 9 predictors. To retain as many participants as possible without overloading the model by including all the predictors in the missing data part of the model, we included the three predictors with the most missing data as dependent variables, which triggered their inclusion in the missing data part of the models on the assumption that they share a multivariate normal distribution with all other dependent variables.

Resident level statistical analyses.

For the latent class analysis models, only the first baseline interview with each of the participants was utilized. For the combined latent class analysis-exit model, the same individuals were included, but their data over time were used to define the exit event.3 We started our approach to the latent class analysis using the standard assumptions. Indicators were assumed to have a multivariate normal distribution within class and the within class covariance matrix was assumed to be diagonal (i.e., uncorrelated indicators) and invariant across the classes.

The model for predictors of latent classes is a multinomial logistic regression. In such models, the choice of a “reference” class (the one all other classes are compared to) is arbitrary; any such choice is as good as any other. Thus, one is free to choose the reference class in a way that best identifies substantively relevant class differences. The reference class choices described further on reflected this goal, and were further tested using nested chi-square tests to determine which was the most parsimonious yet adequate description of the predictor effects

To reduce skewness and avoid potential problems with outliers, we transformed all but two indicators: quality of life and stress. We reversed the scaling of indicators with negative skewness (hope, self-esteem, self-efficacy, psychological sense of community) so that they were positively skewed (similar to wages) and tried a variety of power transformations (e.g., square root) and picked the one that best reduced the positive skewness without inflating kurtosis and resulted in a reasonably normal distribution. Finally, we restored the original scaling of the transformed variable.

All models were estimated using Mplus (version 8.4; Muthen & Muthen, 2017) with corrections to standard errors for correlation of participants within OHs. Our approach to latent class analysis with predictors and outcomes necessarily involves many individual models each with a large amount of primary and auxiliary output so to facilitate the processing of all the Mplus output, MplusAutomation (version 0.8.1) was used for post processing.

House Statistical Analysis

House level latent class analyses were based on house level data at waves 1 and 2 and the latent class analysis indicators included average network density across waves 1 and 2 of friendship, advice seeking and willingness to loan; House president ratings at wave 1 of 1) house involvement in chapter activities, 2) house savings, 3) house new resident selectivity, and 4) typical house resident financial status.

To simplify interpretation of latent class profiles, the indicators were standardized prior to latent class analysis to means of zero and standard deviations of 1. After model fitting, class mean profiles were further standardized using the estimated within class covariance matrix so they represent deviations from the overall population mean in residual, within class standard deviation units (similar to Cohen’s d).

Overall statistical analysis

A logistic regression analysis used evictions as the dependent variable with the two house and five resident classes as the independent variables, with sociodemographic variables as control variables. Chi-squares were also used to examine the house class 1 for the 5 resident groups, and house class 2 for the 5 resident groups; and next house class 1 was compared with house class 2 for each of the 5 resident groupings.

Results

Descriptive Statistics

The analysis sample of 602 was 51% male, with a mean age of 37.0 years (SD = 10.5). Participants identified as European-American (78.8%), African American (8.5%), Latinx (10.0%), with all other ethnicities accounting for 2.7% (Asian American, Alaskan Native, American Indian, and Pacific Islander). Missing data on exits and predictors were minimal, never exceeding 5% of the 602 residents for any variable.

As reported in Jason, Stoolmiller, and Light (2021), for the latent class analysis with the eight indicators of social capital, model fit characteristics of the 5-class model include entropy of .82, which suggests good classification accuracy (on a scale of 0–1, with 1 meaning perfect classification and 0 meaning random chance accuracy; See Figure 1). Descriptively, we used the class number “1” to indicate “high” levels on almost all variables, class number “2” to refer to “medium-high” scores on most variables, class number 3 to refer to “high self-esteem” indicating the class with the highest self-esteem scores, class number 4 to refer to those with “medium-low” scores, and class number 5 to refer to “low” describing the lowest scores on most variables. Class 3, while similar to class 2 on most indicators at a medium high level, has an unusually high level of self-esteem, even higher than class 1. In addition, class 5, while generally the lowest on most recovery indicators, had a higher level of abstinence self-efficacy than class 4.

Figure 1.

Figure 1.

Open in a new tab

Model fitted latent class profiles for the 5-class model with model estimated class proportions in the top margin. Because the indicators used in the model are not all identically scaled, the mean profiles shown are for the standardized (to mean of 0 and standard deviation of 1) indicators. Stress has also been reverse scaled.

For the house level latent class analysis, Figure 2 as indicated by Jason, Stoolmiller, Light, and Bobak (2021) shows that all the indicators had at least 1 significant class contrast in the 2- or 3-class model. Entropy was very good for both the 2- and 3-class models, .90 and .94 respectively, on a scale of 0–1 with 1 being perfect class prediction. We selected the two-class model for the current study in order to increase sample size for the analyses. Figure 2 presents the mean profile plot for the 2-class model and there are prominent differences on each indicator. Those houses that are in Class 1 have higher densities of friendship and willingness to loan, as well as higher member selection criteria, more resident affluence, and higher involvement in chapter activities. The only indicator where Class 2 has a higher mean is the density of advice seeking. Prior research has suggested that high levels of advice seeking are related to less positive outcomes (Jason, Guerrero, Salomon-Amend, Light, & Stoolmiller, in press).

Figure 2.

Figure 2.

Open in a new tab

House 2 class solution.

Table 1 presents the findings for evictions for residents of the two latent house and five latent resident classes. Table 2 presents a logistic regression with evictions as the dependent variable and the house and resident latent variables as well as the interaction of house and resident variables as independent variables. A variety of covariates were not significant (gender, race, age, employment, education status, length of time in the recovery home), so they are not included in the model. Results indicated that for the resident latent factor, class 1 was significantly different than the other four classes but there was not a significant interaction or house effects. Power issues probably made it difficult to find the interaction and house effects.

Table 1.

Evictions by House and Resident Class

Resident Class
1 2 3 4 5
House Class
1 9% 14% 36% 25% 15%
2 13% 40% 43% 46% 40%
Open in a new tab

Table 2.

House, Resident and Interaction Latent Variables in Logistic Regression

Variables in the Equation
B S.E. Wald df Sig. Exp(B)
Step 1a Two House Type * 5
Resident Types		 3.143 4 .534
Two House Type(1) by 5
Resident Types(1)		 −.988 .827 1.429 1 .232 .372
Two House Type(1) by 5
Resident Types(2)		 .124 1.015 .015 1 .903 1.132
Two House Type(1) by 5
Resident Types(3)		 −.531 .822 .417 1 .519 .588
Two House Type(1) by 5
Resident Types(4)		 −.882 1.116 .624 1 .429 .414
5 Resident Types 9.682 4 .046
5 Resident Types(1) 1.512 .571 7.013 1 .008 4.535
5 Resident Types(2) 1.644 .603 7.427 1 .006 5.175
5 Resident Types(3) 1.739 .564 9.497 1 .002 5.689
5 Resident Types(4) 1.504 .616 5.959 1 .015 4.500
Two House Type(1) −.418 .749 .311 1 .577 .659
Constant −1.910 .536 12.703 1 .000 .148
Open in a new tab
a.

Variable(s) entered on step 1: Two House Type * 5 Resident Types, 5 Resident Types, Two House Type.

We approached the next set of analyses with chi-squares, and first focused on eviction rate differences between class 1 houses for the five different patient types. Class 1 houses represent those with high levels of friendship and loaning relationships occurring and lower levels of advice seeking. Other positive characteristics of class 1 houses were in the president’s ratings to selectivity, affluence, and participation in outside chapter activities. For class 1 houses, there was not a significant evictions difference between participant types, x2(4, N=233) = 8.22, p = .08. For those who resided in well organized and facilitating houses, the five different types of people (classed based on aspects of their recovery factor) did well (evictions from 9 to 15%), except for type 3 that had excessive self-esteem (with a directionally higher eviction rate of 36%) and type 4 that was low on most person attributes (eviction rate of 25%). Even though class 5 was lower on most attributes, these individuals had higher scores on abstinence self-efficacy. There were no overall significant differences in negative exits for these conditions.

In contrast, Table 1 indicates there was a significant difference in eviction rates for those living in class 2 type houses, x2(4, N=265) = 11.47, p = .02. In general, participant classes 2–5 had high eviction rates of 40 to 46%, whereas residents from class 1, had low eviction rates (13%).

In addition, we also examined differences for the two types of houses for each of the resident types. There were significant differences for class 2 [x2(1, N=206) = 17.43, p < .01] and class 4 [x2(1, N=194) = 8.13, p <.01] but no significant differences between the two classes of houses for resident class 1 [x2(1, N=76) = .31, p = .57], class 3 [x2(1, N=64) = .19, p = .67], and class 5 [x2(1, N=58) = 2.71, p =.10].

Discussion

The field of Community Psychology focuses on both understanding individuals and their social and cultural contexts, and the current study provides a concrete demonstration of this proposition, as we found that certain individual social capital characteristics might be protective even in less-than-optimal recovery settings, and settings that offer more facilitating relationships might be able to overcome limitations in many with vulnerable social capital indicators. In a sense, both individual and setting characteristics are needed to help us learn about the process of adaptation (Kelly, 2006). For in the current study, even in class 2 houses which had less optimal friendship and resource sharing characteristics, class 1 residents who had higher recovery capital indicators such as self-esteem and self-efficacy, were able to have relatively low eviction rates. In addition, when participants within a range of recovery social capital indicators were living in class 1 houses with more facilitating social relationship properties, they tended to evidence low eviction rates.

Eviction rates tended to be high within class 2 houses, and these houses had significantly lower levels of all indicators except advice seeking. It is very likely that houses with high levels of advice seeking might be an indicator of more stress being experienced, and other research has found high advice seeking to be associated with relapse (Jason, Guerrero, Salomon-Amend, Light, & Stoolmiller, in press). However, one group of residents, those from participant class 1, evidenced relatively low relapse rate even in these less-than-optimal housing settings. Class 1 residents had higher social capital scores on most indicators, so the overall recovery capital resources of this group of residents was able to help overcome the limitations of the class 1 houses. These residents might have had the capacities, inner resources, and resilience to do well even in less-than-optimal settings.

As indicated above, most residents living in more optimal class 1 houses had relatively good outcomes in terms of low eviction rates. However, residents from latent class 3 had directionally higher relapse rates, and they were the group with the highest self-esteem scores, whereas class 3 resident scores on the other social capital indicators were only in the moderate range. It is very possible that residents who have elevated self-esteem in relationship to other social capital indicators might be more at risk of evictions, as they might have a sense of false sense of invulnerability of relapse. Residents from class 4 also had higher eviction rates; this could be due to having the second to lowest scores on the social capital indicators. However, class 5 residents had even lower recovery capital scores, but managed to have low eviction rates. In the prior study (Jason, Stoolmiller, & Light, 2021) this class had lower employment level, higher house poverty level, higher education level, and female gender. Yet, resident class 5 participants had higher scores than class 4 participants on abstinence self-efficacy indicator, which might have represented a protective factor.

Our article clearly overviews the individual and OH dynamics that impact on-going, successful recovery. The multiple levels of analysis and tools were used to explore individual and community capital in order to establish a platform to better understand recovery efficacy. In a sense, the OH is community capital and each individual in the house has to learn to navigate the capital within their house. Our study investigated important questions such as: Is an unstable house a barrier to recovery capital? What allows residents the ability to survive and even thrive in their recovery while in an OH. What allows this level of engagement to happen?

The major limitation in this study was the low sample size of recovery homes, and there were a limited number of individuals that were available within the cells in Table 1. Future studies might expand the number of recovery homes in order to evaluate whether similar findings emerge. Also, the outcome variable was eviction rates, and more long-term data is needed regarding what occurred to residents following their leaving these recovery home settings. Finally, the results are specific to the recovery homes in three states, and also to a particular self-governing model of recovery homes, and therefore the findings might not generalize to other geographic locations or other more staff managed recovery homes. Certainly our findings need to be replicated in other treatment and recovery settings. Finally, our study only examined dyadic relationship and did not consider the effect of social identity and particularly social identification with the house rather with specific housemates.

In summary, the results of this study indicate that both individual and house characteristics matter in how well individuals adapt to recovery homes. Placing those at risk on recovery capital characteristics into houses that have less than optimal relationship and organizational domains increases the chances of eviction from these settings. On the other hand, placing similar vulnerable individuals in recovery in more optimal settings, might increase chances of successful stays in these homes. Just as important, those who vulnerable seem to have generally more positive outcomes when living in the more optimal and supportive settings. These findings suggest that investigators might profit from finding ways to assess both individuals and settings, so that we can better match individuals into settings to increase positive recovery outcomes. This study also points to the need to identify risky social recovery settings and find ways to enhance the overall relational qualities by providing either more instructions, training, or placement of those with more recovery time or leadership skills in these houses.

Acknowledgments

The authors appreciate the financial support from the National Institute on Alcohol Abuse and Alcoholism (grant number AA022763). The authors appreciate the social network help of Ed Stevens. We also acknowledge the help of several members of the Oxford House organization, and in particular Paul Molloy, Alex Snowden, Casey Longan, and Howard Wilkins.

Footnotes

1

One house dropped out completely but another was added after wave 1 bring the total to 43 houses. However, 42 is accurate in the sense that only 42 houses have two or more waves of information available on their residents and a single data point carries no information about change or the probability of an eventual exit.

2

34 entered an OH for the first time at wave 7 and we were not able to determine whether they exited given that the last data collection occurred during wave 7; 15 were forced out because their OH closed before wave 7 and before they filled out a second survey; 15 were missing their reason for leaving and had only filled out one survey before leaving.

3

Six participants were observed to leave the recovery homes during the course of the 2 year study but their reason for leaving could not be determined. We right censored these individuals at their penultimate survey as they had at least one additional wave of data beyond their baseline survey. In addition, 34 residents had more than one OH exit, either from two different OH’s or from the same OH. To avoid greatly complicating the exit model for a very small number of individuals, we only included their first exit in all analyses.

Contributor Information

Leonard A. Jason, DePaul University

Ted Bobak, DePaul University.

Mohammed Islam, DePaul University.

Mayra Guerrero, DePaul University.

John M. Light, Oregon Research Institute

Mike Stoolmiller, Oregon Research Institute.

References

  1. Beasley CR, & Jason LA (2015). Engagement & disengagement in mutual-help addiction recovery housing: A test of affective events theory. American Journal of Community Psychology, 55, 347–358. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Best D, & Laudet A. (2010). The potential of recovery capital. London: RSA. [Google Scholar]
  3. Cacciola JS, Alterman AI, McLellan AT, Lin Y-T, & Lynch KG (2007). Initial evidence for the reliability and validity of a “Lite” version of the Addiction Severity Index. Drug and Alcohol Dependence, 87, 297–302. [DOI] [PubMed] [Google Scholar]
  4. Campbell KL, Ash S, & Bauer JD (2008). The impact of nutrition intervention on quality of life in pre-dialysis chronic kidney disease patients. Clinical nutrition, 27(4), 537–544. [DOI] [PubMed] [Google Scholar]
  5. Cloud W, & Granfield R. (2008). Conceptualizing recovery capital: Expansion of a theoretical construct. Substance Use & Misuse, 43(12–13), 1971–1986. [DOI] [PubMed] [Google Scholar]
  6. Cohen S, Kamarck T, & Mermelstein R. (1983). A global measure of perceived stress. Journal of Health and Social Behavior, 24, 385–396. [PubMed] [Google Scholar]
  7. Cohen S, & Wills TA (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310–357. [PubMed] [Google Scholar]
  8. Garcia-Rea EA, & LePage JP (2010). Reliability and validity of the World Health Organization quality of life: Brief version (WHOQOL-BREF) in a homeless substance dependent veteran population. Social Indicators Research, 99(2), 333–340. [DOI] [PubMed] [Google Scholar]
  9. Jason LA, Guerrero M, Lynch G, Stevens E, Salomon-Amend M, & Light JN (2020). Recovery home networks as social capital. Journal of Community Psychology, 48, 645–657. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Jason LA, Guerrero M, Salomon-Amend M, Light JN & Stoolmiller M (in press). Personal and environmental social capital predictors of relapse following departure from recovery homes. Drugs: Education, Prevention & Policy. Published online Feb. 28, 2021. [DOI] [PMC free article] [PubMed]
  11. Jason LA, Guerrero M, Salomon-Amend M, Stevens E, Light JN & Stoolmiller M. (2021). Context matters: Home-level but not individual-level recovery social capital predict residents’ relapse. American Journal of Community Psychology, 67(3–4), 392–404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Jason LA, Stevens E, & Ram D. (2015). Development of a three-factor psychological sense of community scale. Journal of Community Psychology, 43(8), 973–985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Jason LA, Stoolmiller M, & Light JN (in press). Latent Profile Analysis in recovery homes: A single quantitative dimension captures most but not all of the important details of the recovery process. Substance Abuse. [DOI] [PMC free article] [PubMed]
  14. Jason LA, Stoolmiller M, Light JN, & Bobak T. (2021). House level latent classes as predictors of recovery and evictions. Paper being reviewed by Journal of the Society for Social Work and Research. [DOI] [PMC free article] [PubMed]
  15. Jason LA, Wiedbusch E, Bobak T, & Taullahu D. (2020). Estimating the number of substance use disorder recovery homes in the United States. Alcoholism Treatment Quarterly, 38(4), 506–514. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kelly JG (1979). Adolescent boys in high school: A psychological study of coping and adaptation. Hillsdale, N.J: L. Erlbaum Associates. [Google Scholar]
  17. Kelly JG (2006). Being ecological. An expedition into community psychology. New York, NY: Oxford University Press. [Google Scholar]
  18. Kelly JF, Hoeppner B, Stout RL, & Pagano M. (2012). Determining the relative importance of the mechanisms of behavior change within Alcoholics Anonymous: A multiple mediator analysis. Addiction, 107(2), 289–299. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Latané B. (1981). The psychology of social impact. American Psychologist, 36, 343–356. [Google Scholar]
  20. McLellan AT (2002). Have we evaluated addiction treatment correctly? Implications from a chronic care perspective. Addiction, 97(3), 249–252. [DOI] [PubMed] [Google Scholar]
  21. Muthén LK, & Muthén BO (2017). Mplus user’s guide. Los Angeles, CA: Author. [Google Scholar]
  22. Patterson MS, Russell AM, Nelon JL, Barry AE, & Lanning BA (2020) Using Social Network Analysis to understand sobriety among a campus recovery community. Journal of Student Affairs Research and Practice
  23. Polcin DL, & Henderson DM (2008). A clean and sober place to live: Philosophy, structure, and purported therapeutic factors in sober living houses. Journal of Psychoactive Drugs, 40(2), 153–159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Reise SP, Ventura J, Nuechterlein KH & Kim KH (2005). An illustration of multilevel factor analysis. Journal of Personality Assessment, 84(2), 126–136. [DOI] [PubMed] [Google Scholar]
  25. Rosenberg M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press. [Google Scholar]
  26. SAMHSA (2012). SAMHSA’s working definition of recovery: 10 Guiding Principles of Recovery. Rockville, MD: Substance Abuse and Mental Health Services Administration, Center for Substance Abuse Treatment. Available at: https://store.samhsa.gov/system/files/pep12-recdef.pdf [Google Scholar]
  27. Sklar SM, Annis HM, & Turner NE (1999). Group comparisons of coping self-efficacy between alcohol and cocaine abusers seeking treatment. Psychology of Addictive Behaviors, 13, 123–133. [Google Scholar]
  28. Snyder CR, Harris C, Anderson JR, Holleran SA, Irving LM, Sigmon ST, … & Harney P. (1991). The will and the ways: development and validation of an individual-differences measure of hope. Journal of Personality and Social Psychology, 60(4), 570. [DOI] [PubMed] [Google Scholar]
  29. Sterling R, Slusher C, & Weinstein S. (2008). Measuring recovery capital and determining its relationship to outcome in an alcohol dependent sample. The American Journal of Drug and Alcohol Abuse, 34, 603–610. [DOI] [PubMed] [Google Scholar]
  30. Stevens EB, Buchannan B, Ferrari JR, Jason LA, & Ram D. (2014). An investigation of hope and context. Journal of Community Psychology, 42(8), 937–946. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. The World Health Organization Group (1998). The World Health Organization quality of life assessment (WHOQOL): development and general psychometric properties. Social Science & Medicine, 46(12), 1569–1585. [DOI] [PubMed] [Google Scholar]
  32. Tracy EM, Kim H, Brown S, Min MO, Jun MK, & McCarty C. (2012). Substance abuse treatment stage and personal networks of women in substance abuse treatment. Journal of the Society for Social Work and Research, 3(2), 65–79. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES