Predictors of housing instability and stability among Housing First participants: A 24-month study

Roebuck, Maryann; Agha, Ayda; Nelson, Geoffrey; Distasio, Jino; Ecker, John; Hwang, Stephen W.; Latimer, Éric; Purcell, Sarah; Somers, Julian M.; Tsemberis, Sam; Aubry, Tim

doi:10.1080/10530789.2023.2174565

Cited by 6 publications

(5 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our findings highlight limitations in the current design of homelessness algorithms. Historically, computational homelessness research often used GLMs to identify common characteristics (or types) of unhoused individuals [27,47,50,53,97,107,141]. Our findings showed that this popular methodology has been extended to designing street-level homelessness algorithms due to their explainability.…”

Section: Inferential Statisticsmentioning

confidence: 84%

See 1 more Smart Citation

A Human-Centered Review of Algorithms in Homelessness Research

Moon,

Guha

2024

Proceedings of the CHI Conference on Human Factors in Computing Systems

View full text Add to dashboard Cite

Homelessness is a humanitarian challenge affecting an estimated 1.6 billion people worldwide. In the face of rising homeless populations in developed nations and a strain on social services, government agencies are increasingly adopting data-driven models to determine one's risk of experiencing homelessness and assigning scarce resources to those in need. We conducted a systematic literature review of 57 papers to understand the evolution of these decision-making algorithms. We investigated trends in computational methods, predictor variables, and target outcomes used to develop the models using a human-centered lens and found that only 9 papers (15.7%) investigated model fairness and bias. We uncovered tensions between explainability and ecological validity wherein predictive risk models (53.4%) unduly focused on reductive explainability while resource allocation models (25.9%) were dependent on unrealistic assumptions and simulated data that are not useful in practice. Further, we discuss research challenges and opportunities for developing human-centered algorithms in this area.CCS Concepts: • Human-centered computing → Human-computer interaction (HCI); Empirical studies in HCI ; • Applied computing → Computing in government.

show abstract

Section: Inferential Statisticsmentioning

confidence: 84%

“…For example, Roebuck et al [107], Shelton et al [117], and Wong et al [141] used regression models to determine risk factors associated with homelessness. Culhane and Kuhn [68] used cluster analysis to identify three types of shelter users (transitional, episodic, and chronic).…”

Section: Related Workmentioning

confidence: 99%

A Human-Centered Review of Algorithms in Homelessness Research

Moon,

Guha

2024

Proceedings of the CHI Conference on Human Factors in Computing Systems

View full text Add to dashboard Cite

show abstract

“…These findings align with recent research that has identified overdose as a serious concern in supportive housing programs [ 23 , 41 ]. More broadly, substance use problem severity has been identified as a risk factor associated with housing instability in PSH and continued connections to people who use substances may present eviction and apartment takeover risks [ 15 , 75 ]. The latter findings highlight the complexity of social networks among people who use substances in PSH, as both potential sources of important support and risk [ 76 , 77 ].…”

Section: Discussionmentioning

confidence: 99%

“…In an early examination of data from a multisite randomized controlled trial of Housing First, an intervention often delivered as PSH, findings were only able to predict 3.8% of the variance in housing instability outcomes after 12 months using sociodemographic, clinical, and housing history variables [ 14 ]. Subsequent analyses from the same trial with a more stringent definition of housing stability and set of predictors produced an improved model, but ultimately yielded the same conclusion: Although certain individual characteristics are risks factors associated with difficulties establishing housing stability, the researchers concluded that it was not possible to accurately predict who would be unsuccessful in Housing First after 24 months [ 15 ]. Studies examining associations between service use and housing stability have also produced relatively small effect sizes [ 16 ].…”

Section: Introductionmentioning

confidence: 99%

Managing high-risk behaviours and challenges to prevent housing loss in permanent supportive housing: a rapid review

Kerman,

Kidd,

Mutschler

et al. 2023

Harm Reduct J

Self Cite

View full text Add to dashboard Cite

Permanent supportive housing is an effective intervention for stably housing most people experiencing homelessness and mental illness who have complex support needs. However, high-risk behaviours and challenges are prevalent among this population and have the potential to seriously harm health and threaten housing tenures. Yet, the research on the relationship between high-risk issues and housing stability in permanent supportive housing has not been previously synthesized. This rapid review aimed to identify the housing-related outcomes of high-risk behaviours and challenges in permanent supportive housing settings, as well as the approaches used by agencies and residents to address them. A range of high-risk behaviours and challenges were examined, including risks to self (overdose, suicide/suicide attempts, non-suicidal self-injury, falls/fall-related injuries), and risks to multiple parties and/or building (fire-setting/arson, hoarding, apartment takeovers, physical/sexual violence, property damage, drug selling, sex trafficking). The search strategy included four components to identify relevant academic and grey literature: (1) searches of MEDLINE, APA PsycINFO, and CINAHL Plus; (2) hand searches of three journals with aims specific to housing and homelessness; (3) website browsing/searching of seven homelessness, supportive housing, and mental health agencies and networks; and (4) Advanced Google searches. A total of 32 articles were eligible and included in the review. Six studies examined the impacts of high-risk behaviours and challenges on housing tenancies, with overdose being identified as a notable cause of death. Twenty-six studies examined approaches and barriers to managing high-risk behaviours and challenges in PSH programs. These were categorized into eight types of approaches: (1) clinical, (2) relational/educational, (3) surveillant, (4) restrictive, (5) strategic, (6) design-based, (7) legal, and (8) self-defence. Consistent across all approaches was a lack of rigorous examination of their effectiveness. Further, some approaches that are legal, restrictive, surveillant, or strategic in nature may be used to promote safety, but may conflict with other program objectives, including housing stability, or resident empowerment and choice. Research priorities were identified to address the key evidence gaps and move toward best practices for preventing and managing high-risk behaviours and challenges in permanent supportive housing.

show abstract

“…The model assesses predisposing factors (e.g., personal characteristics such as age and gender), need factors (e.g., mental and physical health conditions), enabling factors (e.g., employment, housing), and health behaviors (e.g., use of services). Based on Gabrielian et al’s (2016) and Roebuck et al’s (2023) use of the Gelberg–Andersen Model, we defined the health behavior domain as access to health services, specifically, access to services in the study agency (e.g., program dose, contacts, program completion, and the use of other agency programs). All the factors in Gelberg–Andersen’s model are posited to influence health and health service outcomes (Gabrielian et al, 2016; Gelberg et al, 2000).…”

mentioning

confidence: 99%

Predictors of length of time in service: Characteristics of people in intensive case management for longer than 5 years.

Roebuck,

Bridger,

Magny

et al. 2024

Psychiatric Rehabilitation Journal

View full text Add to dashboard Cite

Predictors of housing instability and stability among Housing First participants: A 24-month study

Cited by 6 publications

References 33 publications

A Human-Centered Review of Algorithms in Homelessness Research

A Human-Centered Review of Algorithms in Homelessness Research

Managing high-risk behaviours and challenges to prevent housing loss in permanent supportive housing: a rapid review

Predictors of length of time in service: Characteristics of people in intensive case management for longer than 5 years.

Contact Info

Product

Resources

About