This section summarizes findings from studies, including peer-reviewed and “grey” literature (reports and non peer-reviewed programme evaluations), addressing specific aspects of shelters and shelter-related services.  Where possible, systematic evidence reviews are cited. Most individual studies use non-comparative, cross-sectional and/or case-series designs, which are often characterized by small sample sizes, high attrition rates and the use of inadequate and/or unvalidated measures. Caution should be used in interpreting their results and in determining applicability to other contexts and populations.

There has been a significant rise in the number of evaluations on shelters and related services over the past 20 years, in large part as a result of increased calls for accountability and demonstration of value-for-money across the social service sector. While framed as ‘shelter evaluations’, the majority of studies actually examine specific programmes provided in or by shelters and not the overall effectiveness of shelters themselves in reducing women’s exposure to violence, or improving their quality of life, among other important outcomes. Most of the studies have been conducted in higher-resourced countries, particularly within Europe and North America (Sullivan et al., 2008).

As most evaluations are observational and non-comparative in design, the evidence base regarding the effectiveness of shelter services is small and generally weak, and is limited by ongoing deficits in research and lack of rigorous evaluation (Wathen & MacMillan, 2003; Ramsay et al., 2005; Ramsay et al., 2009).

Strong qualitative research has been conducted to examine women’s experiences of shelter services in cases of domestic violence (e.g. Tutty et al., 1999; Haj-Yahia & Cohen, 2009). Cross-sectional and survey studies, mainly from the United States, have found that shelter programmes are among the most supportive and effective resources for those with abusive partners, although there is limited data available for shelter responses to other forms of violence (e.g. Bowker & Maurer, 1985; Cannon & Sparks, 1989; Gordon, 1996; Ham-Rowbottom et al., 2005; Itzhaky & Porat, 2005; Lyon & Sullivan, 2007).

There is a pressing need for rigorous, large-scale studies of shelters and their services which employ comparative designs and appropriate/relevant outcome measures to evaluate the benefits and potential harms of providing shelter services to abused women and children. These studies are challenging to conduct, particularly with significant ethical concerns in engaging survivors for such research (Garza, 2002; Tutty et al, 2006). Studies may address these challenges by employing creative designs (especially in constructing comparison groups), prioritizing women’s safety and well-being, and focusing on a range of service models and outcome measures (i.e. comparing basic shelter services to enhanced, case-coordinated services to ensure a minimum standard of care is provided).

Future research design efforts may benefit from multi-level modeling and complex systems approaches (Sullivan et al., 2008; Javdani, et. al, 2011).

See Outcome evaluation strategies for domestic violence service programs receiving FVPSA funding: A practical guide (Lyon, E., & Sullivan, C. M. 2007), a practical guide to conducting these kinds of evaluations based on the context in the United States. Available in English.