Psychological Services · 2017 · 19 citations · 48 references
Public HealthUrban HealthHealth Services ResearchHomelessnessHousingCauses Of HomelessnessChronic HomelessnessPermanent Supportive HousingRisk Factors For HomelessnessSupported HousingSupportive HousingCommunity HousingHomeless Service Utilization DataHomelessness Chronicity MeasurementMental Health Service Utilization AnalyticsPathways Out Of HomelessnessSupported Housing ModelsSupportive Services Integration
Permanent, community-based housing with supportive services ("supported housing") has numerous favorable outcomes for homeless-experienced consumers. Little is known, however, about consumers who attain but subsequently lose their supported housing. Using mixed methods, we compared persons who retained their supported housing for at least 1 year ("stayers") with those who lost their supported housing within 1 year of move-in ("exiters"). Among persons housed through the VA Supported Housing (VASH) program at the VA Greater Los Angeles between 2011 and 2012, we queried VA homeless registry data to identify stayers (n = 1,558) and exiters (n = 85). We reviewed the medical records of 85 randomly selected stayers and all 85 exiters to compare demographics, homelessness chronicity, era of service, income, presence or absence of a serious mental illness, and health service utilization. From this subsample, we purposively selected 20 stayers and 20 exiters for semistructured, qualitative interviews, and more detailed medical record review. We also performed qualitative interviews and focus groups with VASH staff/leadership (n = 15). Recursive partitioning identified quantitative variables that best-differentiated stayers from exiters. Thematic analyses were performed on qualitative data. Interrelated factors were associated with exiting supported housing: chronic homelessness; low intrinsic motivation; unmet needs for mental health care, substance abuse treatment, and independent living skills; poor primary care engagement; frequent emergency department use; and recent mental health hospitalizations. These findings suggest the value of clinical interventions that address these factors-for example, motivational interviewing or social skills training-adapted to the setting and context of supported housing. (PsycINFO Database Record
48
A Study of Cross-Validation and Bootstrap for Accuracy Estimation and Model Selection
Ron Kohavi · 1995 · 10.7K citations · Full text
Sampling for qualitative research
Martin Marshall · Family Practice · 1996 · 4.1K citations
Carolin Strobl, James D. Malley, Gerhard Tutz · Psychological Methods · 2009 · 2.6K citations · Full text