Toward evidence-based urban design

By Tim Stonor, managing director and Chris Stutz, associate director of Space Syntax Ltd.

We have all become aware of the Government's desire to deliver sustainable communities in the UK 'communities that can stand on their own feet and adapt to the changing demands of modern life'.

This aspiration has been recently evinced in the ODPM's Communities Plan and the Thames Gateway and Northern Way initiatives, and it is also the topic of the Summit 2005 conference planned for next year. Some commentators may debate the philosophy of sustainability, but planners and designers must get on with the process of delivering the Government's vision. What some may not know is that the Government has specified the way this is to be done. From the 1999 White Paper Modernising Government through to more recent guidance issued by the Prime Minister's Strategy Unit , the Government has made it clear that it expects a step change in the way government services are delivered through better use of evidence-based practice, an approach considered essential for meeting long term goals. Others have dealt with the implications of this for urban policy , but little attention has thus far been paid to what it means for the related topic of urban design. We hope this article begins to fill the gap.

The roots of evidence-based practice: medicine We are all familiar with the everyday concept of evidence and its use in a wide range of disciplines. But the phrases 'evidence-based practice' and 'evidence bases' were inspired by a specific approach in medicine known as 'evidence-based medicine' or 'EBM'. EBM developed out of the work of British epidemiologist and medical researcher Archie Cochrane in the early 1970s. It is frequently defined as 'the conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients.' The concept of evidence in EBM is more nuanced than our everyday understanding of the term.

In EBM, evidence does not simply prove or fail to prove the effect of a medical intervention. Instead, specific pieces of evidence are assigned levels of validity, which means they are considered better or worse guides for treatment decisions depending on the type of research from which they arose. In general, the findings of a single primary research study are considered to be less reliable than those substantiated by a systematic review of a number of primary research studies. As for primary research, randomised experiments (where participants are randomly placed in a control group and given placebo measures) are preferable to quasi-experiments (where membership in the control group is not random).

However, evidence from any kind of experiment takes precedence over observation, and observation takes precedence over personal experience. The EBM approach has been applied across a wide range of medical disciplines. Of particular interest to planners and urban designers, it has even applied to the design of hospitals. Evidence-based healthcare design, as it is called in this manifestation, has typically focused on surveying research into the influence on patient recovery times of such factors as lighting, art, windows, and plants and nature.

While some see EBM as 'a paradigm shift, from medical practice based on the accumulation of clinical observation, expertise, and experience to one characterised by a systematic search for rigorous and relevant scientific evidence', others claim that the aim of EBM is to integrate the best, latest and most robust research on treatment options with the proficiency and judgement of individual clinicians: the former helps prevent treatment methods from becoming out-of-date, and the latter ensures that treatments are appropriate for individual patients.

Transferring the model: challenges and modifications As we might expect, the wholesale transfer of the model into urban design or any field treating social problems is a far more complicated matter. For one thing, the processes that operate on cities are arguably more complex and less well understood than those that operate within the human body. Secondly, controlled urban design experiments are much more difficult to conduct than controlled medical experiments. Furthermore, the existing research base in urban design is harder to access comprehensively than its counterpart in medicine, so it is more difficult to perform high-level comparative reviews of research to determine if evidence can be substantiated by a number of studies. Of necessity, then, the Government has set a less rigorous standard for evidence to be used in delivering policies and programmes.

The Prime Minister's Strategy Unit lists the following as typical evidence-generating activities : analysing key patterns in sectoral data analysing public attitudes, behaviours and expectations identifying international best practice examples which can provide some guide to potential futures for the UK ·developing hypotheses about trends and causal links, and testing these hypotheses against available data. However, there are additional complications even with this accommodation. It does not specifically address the institutional and cognitive barriers that planners and designers may face in doing or using academic research. This increases the risk that evidence will be misunderstood or misapplied, leading to, in the words of one commentator, a process of 'endarkenment'. Thus, part of bringing evidence-based thinking in urban design must be the building of research and interpretive skills. Some guidance on structuring this process comes from practitioners in evidence-based healthcare design.

Architect D. Kirk Hamilton proposes a four-level model of evidence-based practice. At level one, practitioners familiarise themselves with the literature of their field and try to incorporate relevant evidence into their work. At level two, they begin to hypothesise about measurable outcomes of specific interventions and then test these hypotheses after the interventions have been made. Level-three practioners begin to share their results publicly in the trade and popular press. Those at level four submit their results to the rigours of peer-reviewed journals, and they may even collaborate with academic researchers. But even a well-trained army of researcher-designers will have a final hurdle to overcome. All of the issues discussed above are related to and exacerbated by the issue of context. Every city is different, and every district or neighbourhood within an individual city is different. To be sure, each individual human being is different from all others in some respects. But the vital systems of all humans respiratory, circulatory, digestive are laid out similarly and work in the same way.

This is not the case with cities. And if different humans respond differently to the same treatment despite the similarity of their systems, how much more difficult it must be to prescribe uniform but effective measures for the problems of cities. This begs the question of whether there can ever be useful evidence-based urban design. Toward a process of evidence-based urban design We think there can be, but it will not be a practice of consulting design copybooks. Design codes as we know them should not be our target. The task is somewhat more difficult: to create a flexible process for deciding the best urban design intervention in any given circumstance. To create this process we need to build up a body of primary research on the effects of design interventions. And to do this, we need to examine the kinds of intervention data we need to collect and the approach we need to collect it.

What interventions should the research focus on? We suggest concentrating on aspects of the design product or process commonly assumed to be important for success. In terms of public spaces, for example, one might examine the effects of consultation, lighting, furniture, etc., across a broad spectrum of cases. This is the type of work begun by people such as William H. Whyte and Jan Gehl and advocated in the UK by CABE. How should we conduct this research? The tenets of evidence-based thinking hold that this research should employ rigorous analytic methods that can be replicated, and it should generate results that can be compared.

This will certainly require more partnerships between professionals and academics. In all cases, it is crucial that we employ methods to control for context in the collection of this primary evidence. Part of the difficulty here arises from the fact that any specific design intervention takes place in a variety of contexts: historical, economic, political, etc. However, arguably the most important context for urban design interventions and the one most often misunderstood, unfortunately is the spatial context. By spatial context, we mean things like the pattern of spatial accessibility (continuity and connectivity) within a site and between a site and its surrounding urban areas.

Patterns of spatial accessibility exert a strong, regular and measurable influence on where and how people move through cities and where and how they stop. For example, in studies of well-functioning areas in cities around the world, we have consistently found that more people tend to move along simpler, spatially accessible routes and to avoid layouts that are labyrinthine and spaces that are too cut off. It is important to identify these patterns because urban spaces are successful to the degree that the requirements of their uses for movement and stopping are met although success is more often credited to the historical inspiration for a design, the economic climate, or political or community aspirations. The recognition that spatial context is a crucial variable in evaluating potential design interventions and a deep understanding of the complexities of applying this knowledge in practice is a product of over 25 years of the type of process we are advocating. It demonstrates the kind of rewards that an evidence-based research programme offers, and it has formed the basis of hundreds of successful research and consulting projects all over the world by the Space Syntax Laboratory, University College London, and our company, Space Syntax Limited.

Building an evidence-seeking design culture It is clear, then, there are neither simple answers nor permanent ones. Regardless of the particular interventions studied, it is vital to expand the record of cases to cover as many circumstances as possible. Because cities are so complex and unique, it is especially important to focus our efforts on the creation of a process, one that allows us to understand and forecast the effects of different interventions in different contexts.

And this should point out that just as important as establishing the evidence base itself is creating an evidence-seeking design culture. Only after such a culture emerges, with vigorous debates over rigorously analysed methods and interventions, can we truly hope to reach an evidence-based urban design practice.