What are Best Practices?
There is much interest in identifying and sharing effective program models within the homelessness sector for the purposes of policy, planning and replication/adaptation. People regularly refer to terms such as ‘best practices’ and ‘promising practices’ to describe what they are doing or what they are looking for. But what exactly do these terms mean?
The goal of ‘best practices’ and ‘promising practices’ research is to evaluate an activity that is determined by peers to be highly effective and representative of a really good practice, and/or because it is unique and of special interest. Best and promising practices can include the following types of interventions:
Activity A way of doing particular kinds of work that may have an impact. Examples include: exploring best practices in outreach services, Housing First or Harm Reduction.
Program A group of related activities that is intended to produce outcomes. An example would be an employment and skills training program that helps people find jobs.
Agency A collection of programs and activities within an institutional framework. A homeless drop-in centre offering a meal program, employment training, ID replacement and housing help services is an example of an agency.
Policy A stated principle or rule to guide decisions and courses of action designed to have an impact on homelessness. Policies are established by governments and other organizations. A ‘discharge planning’ policy, for example, is designed to reduce the flow of individuals from corrections into homelessness.
Community Response An organized effort by a community to address homelessness. In some cases, community responses are ad-hoc, fragmented and uncoordinated. In other cases, communities have implemented a systems approach involving an integrated local or regional system aimed at coordinating responses to homelessness. Coordinated and integrated service responses are often referred to as a ‘system of care’.
The majority of evaluation research that exists on homelessness interventions focuses on programs and practices, with very little that identify effective systems or policy responses. In our framework document, we focus on programmatic interventions in the context of systems, policies and agency dynamics. However, the hierarchy of evidence we are proposing can encompass analyses of the full range of interventions to address homelessness. We begin by differentiating ‘best’, ‘promising’ and ‘emerging’ practices.
A Best Practice is an intervention, method or technique that has consistently been proven effective through the most rigorous scientific research (especially conducted by independent researchers) and which has been replicated across several cases or examples. An organized effort by a community to address homelessness. In some cases, community responses are ad-hoc, fragmented and uncoordinated. In other cases, communities have implemented a systems approach involving an integrated local or regional system aimed at coordinating responses to homelessness. Coordinated and integrated service responses are often referred to as a ‘system of care’.
An intervention is considered to be a Promising Practice when there is sufficient evidence to claim that the practice is proven effective at achieving a specific aim or outcome, consistent with the goals and objectives of the activity or program. Ideally, Promising Practices demonstrate their effectiveness through the most rigorous scientific research, however there is not enough generalizable evidence to label them ‘best practices’. They do however hold promise for other organizations and entities that wish to adapt the approaches based on the soundness of the evidence.
Emerging practices are interventions that are new, innovative and which hold promise based on some level of evidence of effectiveness or change that is not research-based and/or sufficient to be deemed a ‘promising’ or ‘best’ practice.
In some cases this is because an intervention is new and there has not been sufficient time to generate convincing results. Nevertheless, information about such interventions is important because it highlights innovation and emerging practices worthy of more rigorous research.
When we speak more generally of ‘best’ or ‘promising’ practices, then, we are attempting to make a determination as to interventions that actually lead to effective and productive results, based on stated goals. However, in order to make this determination, there must be solid evidence. We cannot simply assert that a given approach is outstanding without being able to demonstrate that it has an impact.
The determination of best and promising practices is, in the end, a question of evidence, and may involve conducting research or reviews of the literature, including meta-syntheses and meta-analyses. It may also involve ‘case studies’; that is, detailed investigations of specific examples. In either case, the goal of conducting such research is to make a determination about effectiveness, so as to assist in the dissemination and adaptation of practices that work. There are many programs and reports that claim to demonstrate promising and best practices. However, there is incredible variation in the quality, rigour and amount of evidence that is used to back up these claims. Indeed, some programs or reports may make these claims without any documented evidence to back them up.
Hierarchy of Evidence
What constitutes good evidence? In a number of disciplines (medicine and psychology, for instance) researchers have identified and defined a hierarchy of evidence that can be used to determine the effectiveness and efficacy of an intervention, by assessing the quantity and quality of data available on that intervention.
The hierarchy identifies the best evidence as having demonstrated research rigour and generalizability. The notion of a hierarchy of evidence stems from the health sciences, and reflects the kinds of research that in most cases rely on experimental methods, where greater value is placed on data collected and analyzed under the most controlled conditions possible. As we will see, achieving such high levels of reliability is problematic in assessing homelessness interventions, and it is worth pointing out that qualitative methods not only produce important and valid results, they can answer questions and describe processes that quantitative methods cannot.
We propose a hierarchy of evidence that is applicable to homelessness research and that aligns with established practice, but more effectively reflects the research and methods typically found within the homelessness sector. The following diagram lists the different levels of evidence. At the top are ‘best practices’, while interventions that are identified as good based merely on opinions, reports or news articles, but which have a limited evidentiary base, stand at the other end of the hierarchy
“The hierarchy identifies the best evidence as having demonstrated research rigour and generalizability.”
See what success looks like
In an era of growing interest in developing solutions to homelessness, it is increasingly important to know what works, why it works and for whom it works. While there is a growing body of academic research on the causes and conditions of homelessness, there is very little research that describes effective interventions in a practical way that helps communities learn from and adapt these initiatives to local contexts. Many communities and service providers in the non-profit sector lack effective tools, resources and capacity to engage in rigorous program evaluation or to disseminate knowledge learned in order to assist service providers and program planners elsewhere. The notion of ‘promising practices’ is about communication; that is to say, alerting those working in the field to a strategy that demonstrates positive results.