I have recently thought a lot about evidence-based design and wanted to share some of these thoughts here.

Doing things in an evidence-based way means systematically considering results from rigorous research when taking decisions in any professional practice: medicine, design, management, HR, education, policy and governance. It is meant to improve outcomes: better treatments and care for patients, better designed buildings that suit users, better run organisations, happier staff, better learning and teaching, more successfully run schools.

Evidence-based medicine is probably most well-known. It was one of the first disciplines to consider a different approach to practice – one that brings rigorous research results together with the individual expertise of the practitioner, as argued by Sackett et al (1996) in BMJ. The fact that there is an independent body, the Cochrane Collaboration, providing comparative research results for medical practitioners certainly helps spreading this approach.

Other disciplines jumped on board, among them evidence-based management (made popular by Pfeffer and Sutton). The two authors explain in a nice piece in the NYTimes how managers can use evidence in the workplace (for instance the fact that pay incentives don’t improve performance, or that stable teams produce better results), and then take decisions based on these insights. However, it seems that most managers do not act this way.  Google are mentioned as an example of evidence-based decision taking: they wanted to figure out what made a good boss, and being Google, the so called Project Oxygen engaged in some real data crunching, including the analysis of over 10,000 observations on managers’ behaviours. However, this is still the exception.

Evidence-based governance is not faring much better, as  a piece in the Guardian on evidence-based policies discusses. Four different areas of policy making are highlighted: deciding on the starting time of the school day (to ensure students learn better), deciding whether to turn schools in academies (to improve students’ performance), deciding how to sentence drug addicts (to ensure lower rates of follow-on crime), and deciding how to spend money on foreign aid (to ensure children in developing countries learn better and spend more time at school). I like the article for several reasons: it shows how powerful evidence-based decision making could be; it highlights ways to actually design experiments to test the impact of interventions (based on large samples and random assignment of intervention to one part of the sample, but not to the other part to measure differences in development); and it also discusses why and how things often do not happen this way (governments are too short-sighted, decision-makers are too ignorant or change to quickly, for instance due to elections). Using evidence should be imperative, as the author of the article, Mark Henderson, who just published ‘The Geek Manifesto‘ argues:  “It isn’t unethical or irresponsible to experiment with education or criminal justice. It’s unethical not to.”

So what about evidence-based design (EBD)? In the design world, an evidence-based practice would mean to engage prospective users of a building, for instance an office (all of them, not only management!) and gather systematic information on working cultures, work activities, needs, desires, aspirations. This then forms the baseline for designers to come up with people-centred solutions. EBD is most popular for buildings in healthcare and rarely used in other workplaces, as I’ve argued in a conference paper in 2008 and as an article in Fastcompany in 2009 shows, too.

Similarly to evidence-based policy making, EBD suffers from ignorance of decision-makers: too few architects know about it or use it. It is also difficult to use evidence in design-related decisions, because randomised trials would not work in the same way as in evidence-based policies – the sample sizes in EBD are simply too small. Effectively sample size is 1 – design is mostly tailor-made for a unique organisation and a unique site/location/building. Therefore comparisons are tricky if not impossible and assigning interventions randomly seems a bad idea. Instead, evidence-based design should follow a pre- and post-occupancy setup, i.e. understanding the individuality of an organisation before a design intervention in a study, use results from this to tailor the design to the needs of the organisation and its people, and then test the impact of the design by repeating the same study afterwards, e.g. 6 months after a refurbishment or move.

It becomes obvious why few architects and clients go down this route: it is labour-intensive, requires patience and long-sighted planning, and incurs additional costs. Still it is worth doing to improve the quality and fit-for-purpose of office buildings. Why does this matter? Because a lot of people still complain about their offices. In a workplace study in 2008 in the UK by Gensler, it was found that in the average organisation only 26% of staff were satisfied with their workplace. Clearly, there’s a lot to improve.

Spacelab architects (who I work with) have used this approach in their workplace consultancy creating stunning offices that the people working there love. And this should be the ultimate goal: designing buildings for people, as I’ve argued in my talk for TEDxUCL.

________

Update (24 Oct 2012): Another good article on evidence-based design has appeared in the November 2012 issue of the magazine OnOffice. It explains to the general public what evidence-based design is, how it works, and how it is used by different consultancies in real life projects.