A review of an article on Examining the association between shelter status and homelessness duration, notably on research approaches: for Housing Studies students

A review of an article on Examining the association between shelter status and homelessness duration, notably on research approaches: for Housing Studies students

 

How to describe the article in Harvard reference format?

 Bansal, P., Li, N., Colford, K. & Wu, H., 2026. Examining the association between shelter status and homelessness duration: A case study of Halifax, Canada. Cities, 179, p.107677. https://doi.org/10.1016/j.cities.2026.107677 [Accessed 20 October 2026].

Highlight the main research issues and research objectives, 2 main academic theories used and the main research methods used in this article.

The article's main research issues, objectives, academic theories, and research methods can be summarized as follows:

Main Research Issues:

• The association between shelter status (whether an individual is sheltered or unsheltered) and the duration of homelessness.
• How shelter status moderates the relationship between social and health-related factors and the likelihood of exiting homelessness.
• The role of shelter status in shaping housing pathways and the variation in homelessness duration among individuals.

Research Objectives:

• To examine how shelter status relates to the length of time individuals experience homelessness in Halifax, Canada.
• To investigate whether shelter status influences the impact of individual vulnerabilities, such as immigrant status, victimization, substance use history, and age at first homelessness, on homelessness duration.
• To provide survival analysis insights into exit probabilities from homelessness over time, comparing sheltered and unsheltered populations.

Two Main Academic Theories Used:

1.    Typology of Homelessness Based on Shelter Utilization Patterns: The article employs the framework developed by Kuhn and Culhane (1998), which categorizes homelessness into transitional, episodic, and chronic types based on shelter usage patterns.

2.    Urban Governance and Housing Pathways Theory: The study draws on theories about how urban governance structures, including shelter access and institutional routes, influence individual homelessness trajectories and duration (Benjaminsen, Dyb, & O'Sullivan, 2009).

Main Research Methods:

• Use of a near-complete count and survey data of homeless individuals in Halifax, including both sheltered and unsheltered populations.
• Statistical analysis involving:
• Independent samples t-tests and Cohen's d effect sizes to identify factors associated with homelessness duration.
• Cox proportional hazards models to assess the relationship between shelter status, social and health covariates, and the timing of exit from homelessness.
• Interaction terms in the models to investigate how shelter status modifies associations between vulnerabilities and duration.
• Survival analysis techniques including survival curves to understand exit probabilities over time.

 

What research approaches (reasoning) [inductive/ deductive/ abductive] and research approaches (methods) [qualitative/ quantitative/ mixed methods research] have been used in this study?

The study primarily employs the following research approaches and methods:

Research Reasoning Approach:

• Deductive reasoning: The study tests specific hypotheses concerning the association between shelter status and homelessness duration, as well as the moderating effects of shelter status on social and health-related factors. This is evident from the use of established theories (e.g., typology of homelessness and urban governance) to guide hypothesis formulation and the statistical testing of relationships using t-tests and Cox proportional hazards models.

Research Methods Approach:

• Quantitative research methods: The study relies on quantitative data from a near-complete survey of homeless individuals in Halifax, including numerical measurements such as homelessness duration, shelter status, and categorical social and health variables. The analysis involves statistical testing (t-tests, Cohen's d), survival analysis, and Cox proportional hazards modeling to quantitatively examine associations and interactions.

There is no indication of qualitative or mixed methods research such as interviews, ethnography, or qualitative thematic analysis in the reported methodology. The focus is on quantitatively analyzing survey data to explore and measure relationships, consistent with a deductive and quantitative research paradigm.

 

Describe 2 main research limitations of the study.

Two main research limitations of the study are:

1.    Temporal Ordering and Causality Limitation: The study uses cross-sectional data collected at a single point in time when all respondents were experiencing homelessness. As a result, it measures time since last housed for ongoing homelessness spells rather than completed spells. This design limits causal inference because it cannot determine whether unsheltered status causes longer homelessness duration, whether longer homelessness increases the likelihood of being unsheltered, or whether both are influenced by earlier barriers to shelter access. Additionally, ongoing homelessness spells may be overrepresented, biasing the data toward longer durations.

2.    Measurement and Data Limitations: Shelter status was only measured at the time of the count, preventing assessment of transitions between sheltered and unsheltered status during homelessness. Homelessness duration was self-reported and subject to recall error or approximations, as indicated by clustering around one-, two-, and three-year marks. Furthermore, social and health-related characteristics were treated as static factors measured at one point in time, so the study cannot determine whether these vulnerabilities preceded or resulted from homelessness