Hierarchy Does Not Make Science More Relevant, or Does It?

Dear Editor, Went through with interest the article entitled “Hierarchy of Evidence: Is One Pyramid Sufficient?, published in the July–September 2025 issue of Amrita Journal of Medicine. The author deserves credit for taking up an important issue through this article.[1] To this debate, I would like to add my own concerns that start with the question as to “Is Typifying Research Designs into hierarchies really required and does this make science we conduct more relevant to the common Public for whom the benefit of science itself is intended for.” Categorizing research designs into rigid hierarchies, with randomized controlled trials (RCTs) and meta-analyses sitting at the top, and case–control or cohort, case reports, and pure observational research deemed less valuable or “lower-tier” needs a relook. The claim that categorizing promotes rigor and standardization does inadvertently make science less relevant and less responsive to the public. The hierarchies limit scientific understanding, by favoring development of “science in silo” and “disciplinary limits” through a marginalization of lived experiences, and thereby erodes public trust in research. In a rapidly evolving world as we live in today, research hierarchies have a tendency to oversimplify complex, real-world phenomena. For example, RCTs, while powerful in isolating causal relationships, are often conducted under controlled conditions that do not reflect the nuanced, interconnected challenges that communities face in terms of social behaviors, financial and physical constraints, environmental factors, cultural dynamics, and individual differences. This limits real-world applicability, leading the public to question the relevance of scientific conclusions to their lived experiences. While science which by definition has to be accurate and unbiased, conducting it rigorously may not actually support it being unbiased as claims of scientists to their papers being novel appear not just inflated but non-contextual as well. So whatever the rigor or the hierarchy (claimed to be based on rigor), most scientific work published largely depends, builds, and extends upon previous data rather than happening in a vacuum to the extent that most published research is nothing more than incremental to prior evidence rather than disruptively innovative.[2] Significantly, empirical studies have increasingly suggested that the majority of scientific findings are non-reproducible. Add to this the fact that basic biological research seems to have the highest rates of irreproducibility, in the range of 60%–100%.[3,4] Here, it is also important to talk about evidence creation since the process of the development of hierarchies itself is supposed to be a reflection of its (evidence-creation) robustness. Evidence creation essentially is a process of simplification of the observations we make. But it unfortunately thrives on compromise and consensus as most of this process of simplification is made largely dependent on adding a proposition to it.[3] All further efforts (whatever the design one use) in creating evidence are actually efforts directed at supporting the same (proposition). Those propositions that the consensus finds true (statistically/mathematically, through peer review or otherwise) are accepted as part of the evidence. But the problem is that statistical significance in itself is dependent on the proposition of earlier studies.[4] Therefore, the probability that a research finding is indeed true depends on the prior probability of it being true, making all types of published research vulnerable to bias and inaccuracy and reducing hierarchies to redundancy. Humans through years of trial and error have realized that just because the systems of equations that we use to model our universe are deterministic, does not mean our universe is. Similarly, hierarchies, though useful in modeling the type of research (probably for simplification), do not make research more or less useful to the end users of the research. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest