If A, Then B.

Abstract

Documentation of design intent and a commitment to measurement is a fundamental precept of an evidence-based design process. For several years and in several ways, I have described this need to document design intent as the need to formally state a design hypothesis, in such a way that an outcome measurement will tell you whether your was supported or not. As a longtime practitioner making the transition to academic scholarship, this analogy made sense to me. My co-editor, who is a true research scholar, has reminded me that hypothesis is a research term used in experimental models and might be misapplied in some practice situations. In other words, except for serious design research, the use of a in the evidence-based design world may be a stretch for project assessment.[Image omitted: See PDF.]I would like to explore some of the language associated with measuring results of a completed design and construction project. Perhaps a few simplistic, nonacademic definitions will help my practitioner colleagues.Correlation. In some cases, we simply wish to understand how A is related to The correlation question can be important for designers. We know, for example, that nurse satisfaction and patient satisfaction are highly correlated. nurse satisfaction is high, patient satisfaction will be high. The opposite is also true. patient satisfaction is high, nurse satisfaction will be high. We know there is a correlation, but we do not know why, or which one leads to or causes the other. Both could be influencing the other. Understanding the A: B correlation could influence a designer to use design to enhance A or B in hopes of enhancing both. But a correlation does not imply cause.Causation. In other cases, we may be more interested in whether A leads to B, as in If A, then B. We can consider B an outcome of A. the relationship is not confused by other factors, we might say that A caused When I turn on the light switch, the light fixture is caused to illuminate. It might be more accurate to say the switch causes electricity to be sent to the light fixture, which in turn produces light, or If A, then C. C, then B. This is a chain of causalityAnother possible situation occurs when A causes C and C is only correlated with B, not a cause of This is If A, then C:B. Perhaps the design intervention (A) was to reduce noise that caused lower noise levels (C), which was then correlated with better sleep and higher patient satisfaction (B). The noise reducing design did not cause higher patient satisfaction.The PICO question. I was taught a simple way to imagine an evidence-based design question. A PICO (peek-oh) question identifies the population or place (P) that a design intervention (I) will impact, as compared (C) with the absence of the intervention, and for which you measure an outcome (O). I would, for example, enjoy being able to find funding to apply evidence about lighting found in the literature in a project for the ICU, so I might think of a PICO question to help plan my evidence-based design project. For the hospitalized critical care patients (the population, P), would a design intervention (I) that provided full spectrum lighting mimicking daylight, as compared (C) with traditional fluorescent lighting, produce earlier patient discharges and reduced the use of pain medications by patients (outcomes, O)? The PICO model makes me state the question in a form that ensures that there is the potential for application of existing evidence from the literature and a serious study or measurement of how the intervention worked in my project's setting.A PICO (peek-oh) question identifies the population or place (P) that a design intervention (I) will impact, as compared (C) with the absence of the intervention, and for which you measure an outcome (O).Randomized, controlled trial (RCT)--The gold standard. The RCT is often described as the gold standard for research. …