Evidence hierarchy exists as a means of evaluating the strength of the evidence that has been provided in a study. The highest order of evidence, for example, is a meta-analysis of randomized-controlled trials (RCTs) that have clear results. The lowest are case reports, which would be viewed as anecdotal in nature. One of the distinguishing features of the hierarchy of evidence is that the best studies are those that can be extrapolated to a larger population, while the weakest are ones that generally cannot be extrapolated. This hierarchy was developed to support the growing call for evidence-based practice in health care (Evans, 2003). My project seeks to compare wait times in outpatient centers compared to traditional emergency room settings. For this study to rank high on the hierarchy of evidence, there would need to be more than just a comparison of wait time statistics -- an independent variable would need to be defined. This would be whatever unique feature distinguishes the outpatient centers from traditional emergency room settings. While knowing that there is a difference in wait times, that would only be a cross-sectional study, which is second-lowest on the hierarchy of evidence. More important, it would not be an experimental study,...

An independent variable, therefore, needs to be identified. Working with an independent variable and then testing the outcomes (wait times) would allow the study to move higher up the hierarchy of evidence. How high would depend on the quantitative strength of the evidence.
There are logical holes in simply looking for correlations in a non-experimental study design. The critics of evidence hierarchies offer all manner of logical fallacies in their criticism -- ranging from "I don't think it's intuitive" to "other people don't use them" (Borgerson, 2009) but such critiques are weak -- the hierarchy of evidence exists specifically to demand rigor rather than wishing away rigor on the grounds that all evidence is equal. It is not. The logic of the hierarchy of evidence is sound -- findings that can be replicated are the best evidence available. Other studies with a lower threshold either do not prove causation, or they simply describe a finding. They have less applicability to evidence-based practice. For example, an anecdote can show that something happened once, but for that intervention to…